State and Local
Climate and Energy Program
Assessing the
Multiple Benefits
of Clean Energy
A RESOURCE FOR STATES
U.S. ENVIRONMENTAL PROTECTION AGENCY
EPA-430-R-11-014 REVISED SEPTEMBER 2011
-------
ACKNOWLEDGEMENTS
This document, Assessing the Multiple Benefits of Clean
Energy: A Resource for States, was developed by the
Climate Protection Partnerships Division in EPA's
Office of Atmospheric Programs. Denise Mulholland
managed the overall development of the Resource.
Julie Rosenberg and Kathleen Hogan (now with U.S.
DOE) provided content and editorial support for the
entire document.
The U.S. Environmental Protection Agency (EPA)
would like to acknowledge the many other EPA em-
ployees and consultants whose efforts helped to bring
this extensive product to fruition.
The following EPA contributors provided significant
assistance through their technical and editorial review
of chapters within the Resource:
Jeffrey Brown, Ben DeAngelo, Andrea Denny, Art
Diem, Nikolaas Dietsch, Steve Dunn (now with U.S.
DOE), Neal Fann, Caterina Hatcher, Kathleen Hogan
(now with U.S. DOE), Bryan Hubbell, Dan Loughlin,
Katrina Pielli, Julie Rosenberg, and Eric Smith.
A multi-disciplinary team of energy and environmen-
tal consultants provided research and editorial support,
as well as technical review of chapters within this
Resource. They include: Stratus Consulting (Heidi Ries,
Joanna Pratt, James Lester, Joe Donahue and Leland
Deck); Synapse Energy Economics (Alice Napoleon,
Bill Steinhurst, Max Chang, Kenji Takahashi and
Robert Pagan); Summit Blue (Kevin Cooney and Mike
Bammel); Energy and Environmental Economics,
Inc. (Snuler Price); Demand Research LEG (Marvin
Horowitz); Abt Associates, Inc. (Mike Fisher, Dan
Basoli); and IGF International (Joshua Smith, Juanita
Haydel, Brad Hurley, Mark Lee, Jay Haney, Bansari
Saha and Karl Hausker).
For more information, please contact:
Denise Mulholland
U.S. Environmental Protection Agency
State and Local Climate and Energy Programs
TEL: (202) 343-9274
EMAIL: mulholland.denise@epa.gov
-------
ASSESSING THE MULTIPLE BENEFITS OF CLEAN ENERGY:
A RESOURCE FOR STATES
Table of Contents
PREFACE v
CHAPTER ONE i
Introduction
CHAPTER ONE CONTENTS
1.1 What are the Multiple Benefits of Clean Energy? ... .3
1.1.1 Energy Savings and Renewable Energy Generation: The Foundation for Benefits ....3
1.1.2 Energy System Benefits ....3
1.1.3 Environmental and Health Benefits. ....5
1.1.4 Economic Benefits ....6
1.2 Why Assess the Multiple Benefits of Clean Energy? 8
1.2.1 Demonstrating the Multiple Benefits of Clean Energy.... 8
1.2.2 Designing or Selecting Options that Achieve Greater or Broader Benefits 8
1.2.3 Identifying Opportunities to Use Clean Energy in Other Planning Processes ....9
1.2.4 Building Support for Clean Energy Policies and Program 11
1.3 How do States Assess the Multiple Benefits of Clean Energy? ... 11
References 14
CHAPTER TWO 19
Assessing the Potential Energy Impacts of Clean Energy Initiatives
CHAPTER TWO CONTENTS
2.1 How Do Clean Energy Policies Affect Energy? 20
2.2 How Can States Estimate the Potential Direct Energy Impacts of Clean Energy Policies? .21
2.2.1 STEP 1: Develop a Business-As-Usual Energy Forecast 22
2.2.2 STEP 2: Quantify Implications of Targets and Goals 33
2.2.3 STEP 3: Estimate Potential Direct Energy Impacts 35
2.2.4 STEP 4: Create an Alternative Policy Forecast 44
2.3 Case Studies 45
2.3.1 Texas Building Code 45
2.3.2 Vermont - Energy and Energy Savings Forecasting 47
TABLE OF CONTENTS | Assessing the Multiple Benefits of Clean Energy I
-------
Information Resources ... 49
References 50
CHAPTER THREE si
Assessing the Electric System Benefits of Clean Energy
CHAPTER THREE CONTENTS
3.1 How Clean Energy Can Achieve Electric System Benefits 53
3.1.1 The Structure of the U.S. Energy System 53
3.1.2 Primary and Secondary Benefits of Clean Energy 54
3.2 How States Can Estimate the Electric System Benefits of Clean Energy 56
3.2.1 How to Estimate the Primary Electric System Benefits of Clean Energy Resources 61
3.2.2 How to Estimate the Secondary Electric System Benefits of Clean Energy Resources 77
3.3 Case Studies 83
3.3.1 California Utilities' Energy Efficiency Programs ....83
3.3.2 Energy Efficiency and Distributed Generation in Massachusetts .... 85
Information Resources.... ... 86
References 89
CHAPTER FOUR 93
Assessing the Air Pollution, Greenhouse Gas, Air Quality, and Health Benefits of
Clean Energy Initiatives
CHAPTER FOUR CONTENTS
4.1 How Clean Energy Initiatives Result in Air and Health Benefits 95
4.2 How States Estimate the GHG, Air, and Health Benefits of Clean Energy 98
4.2.1 Step 1: Develop and Project a Baseline Emissions Profile ... 98
4.2.2 Step 2: Quantify Air and GHG Emission Reductions from Clean Energy Measures 107
4.2.3 Step 3: Quantify Air Quality Impacts 115
4.2.4 Step 4: Quantify Human Health and Related Economic Effects of Air Quality Impacts 119
4.3 Case Studies 124
4.3.1 Texas Emissions Reduction Plan (TERP) 124
4.3.2 Wisconsin - Focus on Energy Program 125
Information Resources.... 127
References 130
CHAPTER FIVE 133
Assessing the Economic Benefits of Clean Energy Initiative
CHAPTER FIVE CONTENTS
5.1 How Clean Energy Initiatives Create Macroeconomic Benefits ... 134
5.1.1 What are the Direct Effects of Demand-Side Initiatives 135
5.1.2 What are the Direct Effects of Supply-Side Initiatives? 136
5.1.3 What are the Indirect and Induced Effects of Clean Energy Initiatives 137
TABLE OF CONTENTS | Assessing the Multiple Benefits of Clean Energy II
-------
5.2.1 Step 1: Determine the Method of Analysis and Level of Effort 138
5.2.2 Step 2: Quantify Expenditures and Savings from the Clean Energy Initiative 148
5.2.3 Step 3: Apply the Method to Quantify Macroeconomic Effects 153
5.3 Case Studies 154
5.3.1 New York: Analyzing Macroeconomic Benefits of the Energy Smart Program 154
5.3.2 Illinois: Analyzing the Macroeconomic Benefits of Clean Energy Development 156
Sampling of State Clean Energy Analyses by Type of Analytic Method ....157
Information Resources.... ... 159
References ... 159
APPENDIX A 163
Catalogue of Clean Energy Case Studies Highlighted in the Multiple Benefits Guide
APPENDIX B 185
Tools and Models Referenced in Each Chapter
TABLE OF CONTENTS | Assessing the Multiple Benefits of Clean Energy
-------
This page is intentionally left blank.
-------
State clean energy initiatives can
produce significant savings in fuel
and electricity costs, as well as other
benefits to the electric system, the
environment and public health, and
the economy.
Assessing the Multiple Benefits of Clean Energy: A Re-
source for States helps state energy, environmental, and
economic policy makers identify and quantify the many
benefits of clean energy to support the development and
implementation of cost-effective clean energy initiatives.
This Resource identifies the multiple benefits of clean
energy and explains why they should be quantified
and considered along with costs. It starts by presenting
clear, easy-to-understand background information on
each type of benefit to help non-specialists understand
how the benefits are generated and what can be done to
maximize them. Building on that foundation, the Re-
source describes analytic options that states can explore
as they conduct and review analyses of clean energy ini-
tiatives. It provides a framework for assessing multiple
benefits, presenting detailed information on basic and
more sophisticated approaches along with descriptions
of tools for quantifying each type of benefit. It also in-
cludes many examples of how states have used multiple
benefits approaches, along with additional resources for
more information.
This groundbreaking document is the first to organize
and present a comprehensive review of the multiple
benefits of clean energy, together with an analytical
framework that states can use to assess those benefits
during the development and implementation of clean
energy policies and programs. Please Note: While the
Resource presents the most widely used methods and
tools available to states for assessing the multiple benefits
of policies, it is not exhaustive. The inclusion of a propri-
etary tool in this document does not imply endorsement
by EPA.
I Assessing the Multiple Benefits of Clean Energy V
-------
This page is intentionally left blank.
-------
CHAPTER ONE
Introduction
Across the nation, states are considering and imple-
menting a variety of clean energy (CE) policies and
programs using energy efficiency, renewable energy,
combined heat and power (CHP) and clean distributed
generation (DG) to meet energy goals such as provid-
ing affordable, clean, and reliable energy for their
citizens. These policies and programs offer multiple
benefits through their ability to:
Reduce demand for energy;
Decrease stress on the energy system;
Mitigate climate change, environmental degrada-
tion, and related human health concerns; and
Promote economic development.
By including the broader set of benefits in the cost-
benefit analyses conducted during planning processes,
states get more comprehensive assessments of their
potential CE investments and are:
Demonstrating how clean energy policies and
programs can help achieve multiple state energy,
environmental, and economic benefits in a cost-
effective way;
Designing or selecting clean energy options that
offer greater energy, environmental, and economic
benefits;
Identifying opportunities where clean energy can
be used to support energy system, environmental,
and/or economic development planning strategies
across the state; and
Building support for clean energy policies and
programs.
U
o
Q
CHAPTER ONE
Introduction
CHAPTER TWO
Potential Energy Impacts of Clean Energy
CHAPTER THREE
Electric System Benefits of Clean Energy
CHAPTER FOUR
Air Quality Benefits of Clean Energy
CHAPTER FIVE
Economic Benefits of Clean Energy
APPENDIX A
Catalogue of Clean Energy Case Studies
APPENDIX B
Tools and Models Referenced in Each Chapter
CHAPTER ONE CONTENTS
1.1 What are the Multiple Benefits of Clean Energy? .... 3
1.1.1 Energy Savings and Renewable Energy
Generation: The Foundation for Benefits .... 3
1.1.2 Energy System Benefits .... 3
1.1.3 Environmental and Health Benefits ....5
1.1.4 Economic Benefits ....6
1.2 Why Assess the Multiple Benefits of Clean Energy? 8
1.2.1 Demonstrating the Multiple Benefits of
Clean Energy 8
1.2.2 Designing or Selecting Options that Achieve
Greater or Broader Benefits 8
1.2.3 Identifying Opportunities to Use Clean Energy
in Other Planning Processes 9
1.2.4 Building Support for Clean Energy Policies
and Programs 11
1.3 How do States Assess the Multiple Benefits of
Clean Energy? 11
References ....14
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy
-------
WHAT IS CLEAN ENERGY?
Clean energy includes demand- and supply-side resources
that meet energy demand with less pollution than that created
by conventional, fossil-based generation. Clean energy
resources include:
Energy efficiency (EE) - refers to using less energy to provide
the same or improved level of service to the energy consumer
in an economically efficient way. Energy efficiency measures
include a wide variety of technologies and processes, can be
implemented across all major energy-consuming sectors, and
may affect all energy sources (e.g., natural gas, electricity, etc).
Renewable energy (RE) - energy generated partially or
entirely from non-depleting energy sources for direct end
use or electricity generation. Renewable energy definitions
vary by state, but usually include wind, solar, and geothermal
energy. Some states also consider low-impact or small hydro,
biomass, biogas, and waste-to-energy to be renewable energy
sources. Renewable energy can be generated on site or at a
central station.
Combined heat and power (CHP) - also known as
cogeneration, CHP is a clean, efficient technology that
improves the conversion efficiency of traditional energy
systems by using waste heat from electricity generation to
produce thermal energy for heating or cooling in commercial
or industrial facilities. CHP systems typically achieve 60% to
80% efficiencies, which is significantly higher than those of
conventional power plants and separate steam units (http://
www.epa.gov/chp/).
Clean distributed generation (DG) - refers to small-scale
renewable energy and CHP at the customer or end-use site.
For more information, visit the U.S. Environmental Protection
Agency's (EPA's) State & Local Climate Web site (www.epa.gov/
statelocalclimate) and the ENERGY STARฎ Web site (http://
www.energystar.gov/).
Assessing the Multiple Benefits of Clean Energy: A
Resource for States provides states with a framework
for evaluating the potential costs and benefits of their
clean energy goals, policies, and programs. It shows
state analysts how the prospective costs and benefits
are derived, enabling them to conduct and manage
analyses, review cost and benefit estimates presented to
them, and make recommendations about the clean en-
ergy options the state should explore or the appropriate
evaluation approaches and tools to use. This Resource:
Describes both simple and more sophisticated
methods for assessing these benefits;
Provides guidance on how to choose among
methods;
STATE CLEAN ENERGY POLICIES AND PROGRAMS
States implement many policies and programs to advance
clean energy, including:
"Lead By Example" programs where the state increases
the use of clean energy in its own government operations,
fleets, and facilities;
Regulatory approaches such as renewable or energy
efficiency portfolio standards, appliance standards,
building codes, interconnection standards; and
Funding and incentive programs such as public benefits
funds, tax incentives, grants, and revolving loan funds.
For more information on clean energy polices and programs,
go to:
EPA State & Local Climate Web site, www.epa.gov/
sta telocalclima te/
Clean Energy-Environment Guide to Action: Policies, Best
Practices, and Action Steps for States (U.S. EPA, 2006). www.
epa.gov/statelocalclimate/resources/action-guide.html
State Clean Energy Lead by Example Guide (U.S. EPA, 2009).
www.epa.gov/statelocalclimate/resources/example.html
* Presents examples of how states are conducting
multiple benefits analysis and using it to promote
clean energy within their states; and
Offers a wealth of resources, including links to
analytical tools, guidance, and studies.
While clean energy resources are broad in source and
impact, this Resource focuses on guidance for estimat-
ing impacts on the electricity system from energy ef-
ficiency and other clean energy resources that affect the
power system. This focus is not meant to diminish the
importance of other clean energy resourcesincluding
energy efficiency that reduces demand for both elec-
tricity and fossil fuels, and energy supplies from renew-
ables and more efficient use of fossil fuelsbut reflects
the more complex nature of the analysis required to
estimate impacts on the electric system.
This chapter provides an introduction to assessing the
multiple benefits of clean energy, including:
A description of the multiple benefits of clean
energy that are covered in this Resource, along with
examples of the findings from studies that have
estimated the actual and potential benefits of a
variety of state and regional clean energy initiatives
(Section 1.1).
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 2
-------
A discussion of why it is important for states to
assess the multiple benefits of clean energy (Sec-
tion 1.2).
An overview of the process and approaches
involved in prospectively assessing the multiple
benefits of clean energy (Section 1.3).
The remainder of the document provides much more
detail about estimating potential energy savings of
clean energy (Chapter 2) and about assessing the future
electric system (Chapter 3), environmental (Chapter
4), and economic (Chapter 5) benefits introduced in
this chapter.
1.1 WHAT ARE THE MULTIPLE
BENEFITS OF CLEAN ENERGY?
Clean energy affects the demand for and supply of con-
ventional energy and can result in positive effects on
the energy system, the environment, and the economy.
To quantify these benefits, it is first necessary to under-
stand how they are produced through energy savings
and renewable energy generation.
1.1.1 ENERGY SAVINGS AND RENEWABLE
ENERGY GENERATION: THE FOUNDATION
FOR BENEFITS
Clean energy initiatives reduce energy consumption
from fossil fuel generation in two ways:
Energy efficiency policies and programs lead to
direct reductions in energy consumption, which in
turn reduces generation requirements.
Renewable energy and clean distributed supply
resources increase the amount of energy from
clean (and more efficient) rather than conven-
tional sources.
Demand-side initiatives usually change the end-use
efficiency of energy consumption.
Supply-side initiatives usually change the fuel/generation
mix of energy supply resources.
States have significant experience quantifying the
actual and potential energy impact of clean energy
policies. For example:
A program evaluation of the New York State
Energy Research and Development Authority's
(NYSERDA) New York Energy $martSM Program
estimated the cumulative annual electricity savings
achieved through 2007 at 3,060 GWh from energy
efficiency, distributed generation, and combined
heat and power. The cumulative annual renewable
energy generation through 2007 was 106 GWh
(NYSERDA, 2008). Combined, these resources
are equivalent to about 2 percent of the amount of
electricity generated in New York in 2006.1
Energy savings and renewable energy generation are
important results of state clean energy initiatives and
the basis for estimating many of the other benefits of
clean energy to the energy system, environment and
public health, and the economy. For example:
An energy efficiency assessment study of the
opportunities in the Southwest showed that wide-
spread adoption of cost-effective, commercially
available energy efficiency measures in homes and
businesses would reduce electricity consumption
by 18 percent in 2010 and 33 percent in 2020 with
a $9 billion investment. These energy savings
would avoid $25 billion in annual electricity supply
costs and $2.4 billion in annual natural gas costs
(SWEEP, 2002).
This section briefly describes each type of benefit. It also
provides examples from recent studies that offer esti-
mates of the multiple benefits of state and regional clean
energy programs. A full list of all studies mentioned
is presented in Appendix A, Clean Energy Studies:
Summary of Benefits Analyses and Findings. Additional
information about the different types of clean energy
options available to states is provided in Appendix A.
1.1.2 ENERGY SYSTEM BENEFITS
Clean energy initiativesin combination with demand
response measures2 can help protect electricity
producers and consumers from the costs of adding
l Patterns and Trends: New York State Energy profiles: 1992-2006. New
York State Energy Research Development Authority. January 2008. http://
www.nyserda.org/publications/Patterns%20eif%20Trends%20Final%20
-%20web.pdf.
2 Demand response measures aim to reduce customer energy demand at times
of peak electricity demand to help address system reliability issues; reduce the
need to dispatch higher-cost, less-efficient generating units to meet electricity
demand; and delay the need to construct costly new generating or transmission
and distribution capacity. Demand response programs can include dynamic
pricing/tariffs, price-responsive demand bidding, contractually obligated and
voluntary curtailment, and direct load control/cycling (DRAM, 2005).
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy
-------
CONNECTICUT INCORPORATES MULTIPLE BENEFITS IN
EVALUATION CRITERIA FOR NEW CAPACITY ADDITIONS
In June 2005, Connecticut policymakers enacted Public
Act 05-01, An Act Concerning Energy Independence (EIA),
which authorized the Connecticut Department of Public
Utility Control to launch a competitive procurement process
geared toward motivating new supply-side and demand-side
resources in order to reduce the impact of Federally Mandated
Congestion Charges on Connecticut ratepayers.
As part of the bid evaluation process, each capacity project is
scored based on a multiple benefits weighting system:
A total of 85% of the evaluation score is based on a
benefit-cost analysis of the project.
A total of 15% of the evaluation score is determined
through the assessment of five other criteria with their
associated weights:
Reduced emissions of SO2, NOX, and CO2 - 5%
Use of existing sites and infrastructure - 2.5%
Benefits of fuel diversity - 2.5%
Front-loading of costs - 2.5%
Other benefits (e.g., transmission reliability,
employment effects, benefits of high level efficiency
suchasCHP)-2.5%
For more information, visit Connecticut's RFP website: http://
www.connecticut2006rfp.com/index.php
new capacity to the system and from energy supply
disruptions, volatile energy prices, and other reliability
and security risks. The following four energy system
benefits are usually recognized as important ways for
clean energy initiatives to reduce the overall cost of
electric service over time.
Avoided energy generation or wholesale energy pur-
chases. Clean energy measures can displace energy,
specifically electricity, generated from fossil fuels
(e.g., natural gas, oil, and coal fired power plants).
Savings include avoided fuel costs and reduced
costs for purchased power or transmission service.
Avoided or reduced need for additional power plant
capacity. Clean energy measures can delay or avoid
the need to build or upgrade power plants or re-
duce the size of needed additions.
Avoided or deferred transmission and distribution
(T&D) investments. Clean energy measures, such as
customer-sited renewables and clean DG (includ-
ing CHP), which are sited on or near a constrained
portion of the T&D system can delay or avoid the
Many state-level clean energy analyses currently do not
quantify emission-related health effectsa clear gap in analysis
and understanding.
This gap can be addressed using EPA tools such as COBRA and
BenMAP, described in Chapter 4, Assessing the Air Pollution,
Greenhouse Gas, Air Quality and Health Benefits of Clean
Energy Initiatives.
need to build or upgrade T&D systems or reduce
the size of needed additions.
Avoided energy loss during transmission and distri-
bution (T&D). The delivery of electricity results in
some losses due to the resistance of wires, trans-
formers, and other equipment. For every unit of
energy consumption that a clean energy resource
avoids, it has the potential to reduce the associated
energy loss during delivery of energy to consumers
through the T&D system. Distributed resources
also reduce these losses by virtue of being closer to
the load.
Other energy system benefits that can accrue from
clean energy programs include avoided ancillary
service costs, reductions in wholesale market clearing
prices, increased reliability and power quality, avoided
risks (e.g., risks associated with the long lead-time
investments for conventional generation and from
deferring investments until environmental and climate
change policies are known), and improved fuel and
energy security.
Many state and regional studies have quantified these
benefits. These studies include:
A study of the Million Solar Roofs initiative in
California estimated that the program resulted in
avoided capacity investments of about $7.1 million
from 2007-2016 (Cinnamon et al., 2005).
A study of widespread energy efficiency deploy-
ment in the Southwest (introduced in the previous
section), used the calculated potential energy sav-
ings to estimate avoided capacity investments of
about $10.6 billion by 2020 (SWEEP, 2002).
Analyses also illustrate how clean energy programs can
improve the security, diversity, and overall reliability of
a state's energy system, which remains a critical energy
policy objective in light of the vital link between elec-
tric reliability and economic security.
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 4
-------
CLEAN ENERGY INITIATIVES CAN BENEFIT ECONOMIC
DEVELOPMENT
A 2007 study by the American Solar Energy Society assessed
the renewable energy and energy efficiency market and
developed forecasts of the market's future economic growth.
The study established a baseline of 2006 data describing the
size and scope of the renewable energy and energy efficiency
industry, and forecast the growth of the renewable energy and
energy efficiency industry from this baseline to 2030 under
three different scenarios.
Using this approach, the authors developed a case study for
Ohio, an area hard hit by the loss of manufacturing jobs. In
2006 in Ohio, gross revenues for renewable energy totaled
nearly $800 million and the renewable energy industry created
more than 6,600 jobs, including increased employment among
scientific, technical, professional, and skilled workers. The
analysis concluded that the energy efficiency and renewable
energy industries offer significant development opportunities
in the state. In 2030, the renewable energy industry in Ohio
could generate nearly $18 billion in revenues and 175,000 jobs
annually, and the energy efficiency industry could generate
more than $200 billion in revenues and more than 2 million
jobs annually.
Source: Bezdek, 2007.
The financial implications of the East Coast black-
out in August 2003 help illustrate the importance
of a reliable energy system: the blackout, which
lasted a couple of days and affected about 20
percent of the U.S. population, was estimated to
result in economic losses of $4.5 to $10 billion
(Conaway, 2006).
A study of the energy system benefits of energy
efficiency and renewable energy in New England
from Public Benefits Funds (PBFs) programs and
Renewable Portfolio Standards (RPS) concluded
that based on 2004 forecasts from the Capacity,
Energy, Load and Transmission (CELT) report
from ISO-New Englandregional demand-side
management activities would reduce peak demand
by 1,421 MW from a forecasted peak of 27,267
MW, a reduction of about 5 percent (RAP, 2005).
1.1.3 ENVIRONMENTAL AND HEALTH
BENEFITS
Fossil fuel-based electricity generation is a major
source of air pollutants that pose serious risks to public
health, such as increased respiratory illness from fine-
particle pollution and ground-level ozone. Fossil fuel-
based generation is also a major source of greenhouse
gases (GHGs), such as CO2, which contribute to global
climate change. States concerned about emissions are
turning to clean energy technologies to limit pollution
and improve air quality and public health. The air and
health benefits of clean energy are summarized below.
Reduced criteria air pollutant and GHG emis-
sions. This Resource focuses on two categories of
air emissions from the electricity sector: criteria
air pollutant emissions, and GHG emissions. In
the electricity sector, clean energy resources can
reduce these emissions by displacing fossil fuel
generation.3 Reduced emissions of criteria air
pollutantsozone (O3), carbon monoxide (CO),
nitrogen oxides (NOX), sulfur dioxide (SO2), par-
ticulate matter (PM), and lead (Pb)are linked
directly to changes in air quality and public health
effects.4 State actions to reduce GHG emissions are
tied to reducing the risk of global climate change
and generally focus on reducing emissions of CO2.
Criteria and GHG emission reductions are usually
measured in tons or as a percentage of some base-
line level of emissions.
Improved air quality.5 Reduced emissions of criteria
pollutants lead to fewer unhealthy air quality days
and lower the incidence of public health effects as-
sociated with them. Ambient air concentrations of
criteria pollutants are usually measured in "parts-
per" units such as ppm (parts per million) or in
3 It is important to note that estimating reductions in emissions from clean
energy in the presence of market-based emissions programs, such as a cap and
trade program, is more complicated. In the presence of an emissions cap and
trade program (for example the SO2cap and trade program under Title IV
of the Clean Air Act Amendments), sources affected by the cap scale back the
amount of electricity they generate from affected sources and therefore reduce
overall emissions as a result of clean energy. However, because the program
allows these sources to emit up to the number of allowances they hold, they
may adjust their compliance decisions in a way that allows them to generate
these reduced levels of electricity at a higher emissions rate and reduce compli-
ance costs. The allowance price would in theory be reduced. There are ways to
capture the environmental benefits from clean energy for pollutants' affected
market programs, such as retiring a portion of the allowance associated with
the reduction. See Guidance on SIP Credits for Emissions Reductions from
Electric Sector Energy Efficiency and Renewable Energy. U.S. EPA, Office of
Atmospheric Programs, Augusts, 2004. http://www.epa.gov/ttncaaal/tl/
memoranda/ereseerem^d.pdf
4 In addition to being a major source of criteria air pollutants and green-
house gases, coal-burning power plants are the largest human-caused source
of mercury emissions to the air in the United States, accounting for over 50%
of all domestic human-caused mercury emissions (http://cfpub.epa.gov/eroe/
index.cfm;fuseaction=detail.viewInd&lv=tist.tistByAlpha&r=188199&subt
op=341). This Resource, however, does not address methods to assess hazard-
ous air pollutants, like mercury.
5 Improved air quality represents only one of a broad set of environmental
benefits that may accompany clean energy development. Other potential
benefits include improved water quality and improved aquatic habitat. This
Resource focuses on improved air quality and human health
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 5
-------
mass per volume units such as (ig/m3 (micrograms
per cubic meter).6
Improved public health. Improvements in air qual-
ity can reduce the adverse public health effects
resulting from exposure to air pollution and reduce
the costs of associated public health risks. Public
health effects include premature mortality and
exacerbation of health conditions such as asthma,
respiratory disease, and heart disease.
Studies of the environmental benefits of clean energy
initiatives tend to either focus on specific emission
reduction objectives or analyze the overall emission
reductions of multiple pollutants, including GHGs and
criteria pollutants. Examples of these studies include:
A Texas Emissions Reduction Plan (TERP) analysis
in 2004 assessed the potential for clean energy to
help meet NOX air quality requirements as part of
a State Implementation Plan (SIP) and found that
NOX emissions would be reduced by 824 tons per
year in 2007 and 1,416 tons per year in 2012 (Hab-
erl et al., 2004). Texas NOXemissions from electric-
ity generation were 140,676 tons in 2005, so these
reductions represent 0.5 percent and 1 percent of
2005 emissions, respectively (USEPA, 2007).
A 2007 Wisconsin study measured CO2, SO2, and
NOX emission reductions from the state's Focus on
Energy program and found annual emission dis-
placements of 1,365,755 tons of CO2, 2,350 tons of
SO2, and 1,436 tons of NOX from 2001 through 2007
(Wisconsin, 2007)7 These reductions respectively
represent about 2 percent, 1 percent, and 2.5 percent
of Wisconsin emissions in 2005 (USEPA, 2007).
These and other studies demonstrate that clean energy
initiatives can reduce emissions of both criteria air pol-
lutants and GHGs. States may thus find it valuable to
quantify the full range of emission benefits for policy
support purposes.
Fewer studies have quantified the public health ben-
efits of clean energy initiatives. Methods to translate
emissions reductions into changes in air quality and
associated health benefits can be complicated, and until
recently they have not been as accessible to states as
6 For more information on the National Ambient Air Quality Standards
(NAAQS), see http://www.epa.gov/ttn/naaqs/.
7 Emission reductions were presented in pounds in the Wisconsin report but
converted to short tons to simplify comparisons in this document.
methods to assess emissions benefits. One study that
did report health effects provides some indication of
the magnitude of potential health benefits associated
with policies targeting GHG emissions. This study ana-
lyzed how actions to reduce GHG emissions from fossil
fuel use can also reduce conventional air pollutants in
the United States. It found that NOx-related morbid-
ity and mortality benefits, per ton of carbon reduced,
range from $7.5-$13.2 dollars under different carbon
tax scenarios. In addition, the study reviewed 10 prior
studies that estimated health and visibility benefits on
a "per ton of carbon reduced" basis, finding these ben-
efits to range from $3-$90 per ton of carbon emissions
reduced (Burtraw et al., 2001).
1.1.4 ECONOMIC BENEFITS
Clean energy can create broad and diverse economic
benefits that vary considerably across economic sectors
and over time. Many of the energy system, environmen-
tal, and human health benefits of clean energy described
above yield overall economic benefits to the state.
Key economic benefits include:
Energy Cost Savings. Measures that reduce con-
sumers' demand for energy result in energy cost
savings to consumers.8 Once energy savings are
known, energy cost savings can be estimated by
applying a cost factor (e.g., $/kWh) to the energy
savings estimate. Energy cost savings are typically
reported in total dollars saved.
Human Health Benefits. Clean energy policies
that reduce criteria air pollutants may improve air
quality and avoid illnesses and deaths as described
above. Avoided illnesses result in reductions in sick
days taken by employees, increases in productivity,
and decreases in hospitalizations associated with
upper and lower respiratory illnesses and cardiac
arrest. Avoided deaths of workers can result in
continued economic benefits to the state.
Employment. Clean energy initiatives create
temporary, short-term jobs as well as long-term
jobsboth directly from the clean energy activi-
ties and indirectly via economic multiplier ef-
fects. Employment effects of clean energy can be
expressed by many different indicators, such as the
full-time equivalent (FTE) number of jobs or job-
years created. Because an initiative can generate
8 Measures that reduce energy demand may also result in lost revenues for
energy suppliers, at least in the short term.
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 6
-------
OTHER ECONOMIC BENEFITS TO CONSIDER: REDUCING
NATURAL GAS PRICES THROUGH INCREASED DEPLOYMENT
OF RENEWABLE ENERGY AND ENERGY EFFICIENCY
A recent study by the Lawrence Berkeley National Laboratory
(LBNL) examined several studies of the natural gas consumer
benefits from clean energy programs, and analyzed their results
in the context of economic theory. Most of the studies evaluated
a national or state RPS, or a combined RPS and EE program.
Studies in the LBNL analysis consistently found that "RE and EE
deployment will reduce natural gas demand, thereby putting
downward pressure on gas prices" (Wiser et al., 2005). While
the natural gas price reductions vary considerably from state to
state, the analysis did offer some broad conclusions:
Each 1% reduction in national gas demand is likely to lead
to a long-term average reduction in wellhead gas prices of
0.8% to 2%.
Most of the studies that were reviewed and that evaluated
national RPS proposals, found the present value of natural
gas bill savings from 2003-2020 within the range of
$10 - $40 billion.
Consumers' gas bill savings from development of RE and
EE for electric power generation and consumption are
estimated between $7.50 and $20 for each megawatt hour
(MWh) of electricity produced by RE or saved with EE.
Source: Wiser et al, 2005
both employment gains and losses and because
employment effects are likely to vary over time, it
is important for a comprehensive analysis of clean
energy initiatives to assess not only the quantity
of jobs created (or eliminated), but also the type,
duration, and distribution of jobs across the state's
economic sectors.
Output. Economic output is the dollar value of
production, including all intermediate goods
purchased, and all value added (the contribution of
a sector to the economic output). Output depends
upon consumption in the local economy, state gov-
ernment spending, investment, and exports of the
industries in the state. Clean energy programs can
increase output by stimulating new investments
and spending within a state.
Gross State Product. Gross state product (GSP) is
the sum of value added from all industries in the
state, and is analogous to the national concept of
GDP. GSP is equal to the state's economic output
less intermediate inputs acquired from beyond the
state. Clean energy has the potential to result in
GSP increases.
Income. Income effects from clean energy invest-
ments can be measured using a variety of indica-
tors. Most commonly, income effects are expressed
as a change in personal income or disposable
income. Personal income is the sum of all income
received. Disposable income is the income that is
available for consumers to spend or save; that is,
personal income minus taxes and social security
contributions, plus dividends, rents, and transfer
payments. In both cases, a net increase in income
associated with clean energy initiatives can occur
due to increased employment or wages.
Most economic analyses of clean energy initiatives
report results in terms of effects on income, output,
and employment. In several instances, benefit findings
are summarized in terms of the expected benefit per
dollar invested in a clean energy program or per dollar
of energy savings. These values can vary significantly
depending upon the type of value being estimated
and upon the assumptions used to estimate them.9
Examples of findings on the economic effects of energy
efficiency and renewable energy programs include:
Illustrative findings for income and output
Every $1 spent on concentrated solar power in
California produces $1.40 of additional GSP
(Stoddard et al., 2006).
> Every $1 spent on energy efficiency in Iowa
produces $1.50 of additional disposable in-
come (Weisbrod et al., 1995).
Every $1 million in energy savings in Oregon
produces $1.5 million of additional output and
about $400,000 in additional wages per year
(Grover, 2005).
Illustrative findings for employment effects
Every $1 million of energy efficiency net ben-
efits in Georgia produces 1.6-2.8 jobs (Jensen
and Lounsbury, 2005).
Every $1 million invested in energy efficiency
in Iowa produces 25 job-years, and every
9 It is important to understand how any benefit per dollar spent was gener-
ated. For example, some valuesnet valuesconsider the opportunity cost of
how the investment in clean energy could have otherwise been spent. Others
do not consider this cost and may depict a higher return per dollar invested.
For another example, employment benefits may be measured in job-years,
which can be short-lived, and are not the same as net jobs, which are per-
manent, longer term positions. For more information about how values are
calculated and key questions to consider, see Chapters, Section 5.1.
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 7
-------
$1 million invested in wind produces 2.5 job-
years (Weisbrod et al, 1995).10
Every $1 million invested in wind or PV
produces 5.7 job-years, versus 3.9 job-years for
coal power (Singh and Fehrs, 2001).
1.2 WHY ASSESS THE MULTIPLE
BENEFITS OF CLEAN ENERGY?
States have historically evaluated clean energy policies
based predominantly on their costs and impacts on
energy demand. However, by considering the multiple
energy system, environmental, and economic benefits
of clean energy as they design and select clean energy
policies and programs, states can more fully under-
stand the range of costs and benefits of these potential
actions. As stated earlier, with this multiple benefits
information, states can:
Demonstrate how clean energy policies and
programs can help achieve multiple state energy,
environmental, and economic benefits in a cost-
effective way;
Design or select clean energy options that
maximize energy, environmental, and economic
benefits.
Identify opportunities where clean energy can be
used to support energy system, environmental,
and/or economic development planning strategies
across the state; and
Build support for clean energy policies and
programs.
1.2.1 DEMONSTRATING THE MULTIPLE
BENEFITS OF CLEAN ENERGY
Clean energy policies and programs typically reduce
energy demand or increase generation from clean en-
ergy sources. Policies and programs are pursued based
on an assessment of the costs of the program compared
with the results, typically the energy savings or the new
supply of clean electricity. For some options (e.g., low-
cost energy efficiency measures), cost effectiveness can
10 The difference in employment effects between energy efficiency and renew-
able wind power results primarily from the relatively low labor intensity of
energy sectorsboth renewable and fossil fuelcompared with the economy
as a whole. Conserving energy reduces the energy bills paid by consumers and
businesses, thereby enabling ongoing spending of those energy savings on non-
energy goods, equipment, and services in sectors of the economy that employ
more workers per dollar received.
be easy to demonstrate because the direct, near-term
benefits are recognized through less consumed energy
and lower energy costs. However, other project types
(e.g., renewable technologies, higher-cost energy ef-
ficiency measures) require higher initial capital costs,
and may not result in net savings for many years.
When evaluating these types of options on a cost basis
alone, the savings may not exceed the costs during the
short payback period defined by many investors and
utilities (i.e., high discount rates), limiting interest in
the higher investment options.
Most clean energy options, however, result in addition-
al benefits that are frequently left out of the cost-benefit
equation. This omission understates the benefits of
the programs and can limit the use of clean energy to
address multiple challenges. By developing and shar-
ing information about the multiple benefits of clean
energy, states can help build support for their programs
and encourage other states to implement similar clean
energy programs.
For example, the governor of a state may have set
renewable energy goals that are to be achieved through
the states clean energy programs. The same state may
also have economic development challenges, electricity
congestion, or areas of nonattainment under National
Ambient Air Quality Standards and not realize the ex-
tent to which the clean energy programs implemented
to achieve the renewable energy goals also achieve
these other goals by reducing stress upon the electricity
system, reducing GHGs and air pollution, and achiev-
ing public health benefits. By evaluating the potential
energy, economic, and environmental impacts of a
clean energy program, a state can more fully appreciate
the range of its benefits and better understand its cost-
effectiveness. Demonstrating these findings both with-
in and outside the state will help the state gain needed
buy-in for its clean energy program from state officials,
policy makers, and stakeholders, and encourage other
states to implement similar clean energy programs.
1.2.2 DESIGNING OR SELECTING
OPTIONS THAT ACHIEVE GREATER OR
BROADER BENEFITS
Clean energy policies are typically recommended or
implemented based on their potential to meet a specific
goalusually energy-relatedas set by the state. When
selecting among specific clean energy options, how-
ever, it is important to develop a set of more specific
criteria for determining which options to include in the
state clean energy portfolio. Developing these criteria
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 8
-------
How Many Jobs Can The
Clean Energy Industry
Generate?
The University of California-Berkeley re-
viewed 13 independent reports and devel-
oped a model to examine the job creation
potential of the renewable energy industry.
The study analyzed the employment impli-
cations of three national 20% RPS scenarios
and two scenarios where the generation
required by the RPS is produced instead by
fossil-fuel generation.
The key finding is that the renewable en-
ergy industry generates more jobs than the
fossil-fuel industries per unit of energy de-
livered and per dollar invested (Kammen et
al., 2004). Renewable energy's employment
advantage is driven primarily by the general
shift from mining and related services to
increased manufacturing, construction,
and installation activity. The distinction
between renewable technologies in terms
Comparison of Average Employment from Five Electricity Generation Scenarios
and Fuel Processing
a Construction, Manufacturing. Installation
RPS 1: SEW biomass. RPS 2: 50% biomass, RPS 3: 40% biomass,
1-% ',vind energy. 37% wind energy. 55% ;vind energy.
1% solar PV 3% solar PV E% solar PV
Fossil Fuels as Usual:
E0% coal and E0%
natural gas
Gas Intensive: 100%
natural gas
atM from Karrunan it at.. 2CC4
Scenario
of the number of jobs created in O&M and
fuel processing is less clear and technology
dependent. The graph summarizes these
findings.*
involves balancing priorities and requirements specific
to the states needs and circumstances. Assessment
criteria used by states can involve, for example, energy
savings (e.g., in kWh or dollars), economic costs and
benefits (e.g., as measured by payback periods, life-
cycle costs), environmental impacts (e.g., changes in
GHG and air pollutant emissions), economic develop-
ment (e.g., jobs created or lost), and feasibility (e.g.,
political feasibility, time frame for implementation).
For example, the Vermont State Agency Energy Plan
for State Government stresses the importance of
selecting and implementing its clean energy "lead by
example" activities based on several criteria: reducing
state operating costs through energy savings; reducing
environmental impacts; sustaining existing and creat-
ing new Vermont businesses that develop, produce, or
market environmentally preferable products; and dem-
onstrating the economic benefits of clean energy activi-
ties to other states and the private sector (Vermont,
2005). By evaluating potential clean energy activities
with criteria that cut across the multiple benefits, Ver-
mont is able to select options that facilitate the achieve-
ment of multiple state goals and avoid options that may
impede key priorities.
1.2.3 IDENTIFYING OPPORTUNITIES TO
USE CLEAN ENERGY IN OTHER PLANNING
PROCESSES
Many opportunities exist for states to integrate their
clean energy programs with other state environmental,
energy system, and economic programs. States can also
use the multiple benefits from clean energy programs to
help support and strengthen their environmental, en-
ergy planning, and economic development programs.
Using Clean Energy to Achieve
Environmental Goals
Many states and regions are incorporating clean energy
into their environmental strategies to meet their air
quality and climate change objectives. Quantifying the
multiple benefits of clean energy programs can provide
key data for use in developing the SIPs, GHG emis-
sions reduction plans, and air pollution and/or GHG
emissions cap and trade programs that include clean
energy programs. For example, in 2001, the 77th Texas
Legislature established the Texas Emissions Reduction
Plan (TERP) with the enactment of Senate Bill 5 (SB 5),
and recognized that energy efficiency and renewable
energy measures can make an important contribution
to meeting National Ambient Air Quality Standards
in the state. The 78th Legislature further enhanced the
use of clean energy measures to meet the TERP goals
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy
-------
by requiring the Texas Commission on Environmental
Quality to promote energy efficiency and renewable
energy to meet ambient air quality standards (for more
information about the TERP, see Case Studies in Chap-
ter 4, Assessing the Air Pollution, Greenhouse Gas, Air
Quality, and Health Benefits of Clean Energy Initiatives).
States are relying heavily upon clean energy measures
in their climate change action plans to reduce CO2
emissions from the electric power sector. Other states
or regions are using clean energy to advance reductions
under their SO2 and NOx cap and trade programs. For
example, set-asides or carve-outs reserve a portion
of the total capped allowances to be distributed to
clean energy initiatives. Renewable energy and energy
efficiency programs are also being used as offsets in
cap and trade programs focused on reducing GHG
emissions. For example, the Regional Greenhouse Gas
Initiative (RGGI) has developed an offset program in
which heating oil and natural gas efficiency improve-
ments, landfill gas projects, and projects that reduce
sulfur hexafluoride (SF6) can be used as emission re-
ductions. Additional renewable energy and energy ef-
ficiency programs are expected to qualify in the future.
Using Clean Energy to Achieve Energy
Planning Goals
Many state and regional energy plans include clean
energy activities and goals. States analyze the benefits
of these goals to provide a basis for determining which
clean energy initiatives to include in the plan. States
can also require utilities to develop plans that are con-
sistent with these state goals. Utilities are required to
file either integrated resource plans (IRPs) or portfolio
management strategies with the state public utility
commission, depending upon whether the state has a
regulated or deregulated electric system. These IRPs
or portfolio management strategies often use multiple
benefits analysis in the program evaluation criteria.
For example, California requires consideration of en-
vironmental factors in determining cost-effectiveness
of supply- and demand-side options. Beginning in
2003, California's Energy Action Plan has defined an
environmentally friendly "loading order" of resource
additions to meet the electricity needs: first, energy
efficiency and demand response; second, renewable
energy and distributed generation; and, third, clean
fossil-fueled sources and infrastructure improvements
(CPUC, 2003).
MULTIPLE BENEFITS ANALYSIS IS BEING USED
IN REGIONAL PLANNING
The Conference of New England Governors and Eastern
Canadian Premiers (NEG-ECP) seeks to cost-effectively
coordinate regional policies that reflect and benefit U.S.
states and Canadian provinces. In 2001, it developed a
comprehensive Climate Change Action Plan with the long-term
goal of reducing GHG emissions in the region by 75-85%. At
the 30th annual conference held in May 2006, the Governors
and Premiers enacted Policy Resolution 30-2 to promote
energy efficiency and renewable energy in the region. Much
of the resolution was based on a study that quantified the
multiple benefits of existing and expected energy efficiency
and renewable energy programs in New England.
The study. Electric Energy Efficiency and Renewable Energy In
New England: An Assessmen t of Existing Policies and Prospects
for the Future, estimates that by 2010, the combined effect of
expected energy efficiency and renewable energy deployment
will provide a wide range of benefits that go beyond direct
energy savings, including:
Energy System Benefits: the report finds significant benefits to
energy security including a stabilizing and reducing influence
on the wholesale price of, and demand for, natural gas;
reduced wholesale electricity prices in the regional market;
reduced demand for new facilities in the electric market; and
increased resiliency of the grid.
Environmental Benefits: estimated environmental benefits
include savings of 31.6 million tons of CO2 emissions, 22,000
tons of NOX emissions, and 34,000 tons of SO2 emissions
between 2000 and 2010.
Economic Benefits: energy efficiency and renewable energy
programs are estimated to produce a net positive $6.1 billion
for the New England economy, more than 28,000 job-years,
and $1 billion in wages.
Source: RAP, 2005.
Using Clean Energy to Achieve Economic
Development Goals
Clean energy measures yield economic benefits
that can affect businesses, industry, consumers, and
households. Clean energy can create short-term jobs
during the construction of clean energy facilities as
well as permanent long-term employment. Sustained
investment in clean energy can lead to local jobs in
manufacturing, distribution, retail sales, installation,
auditing and rating, and maintenance of equipment
and technology. Cost-effective clean energy can in-
crease regional economic output and reduce energy
bills. As a result, many states are looking to measure
and promote the employment and other economic
development benefits of clean energy, and to incorpo-
rate these benefits into their economic development
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 10
-------
planning processes. In July 2008, for example, Penn-
sylvania Governor Rendell announced and signed The
Alternative Energy Investment Fund. This fund was
created to invest $665.9 million into alternative energy,
including $237.5 million specifically targeted toward
helping consumers conserve electricity and to manage
higher energy prices, and $428.4 million to spur the
development of alternative energy resources and to cre-
ate at least 10,000 well-paying jobs in these industries
(Pennsylvania, 2008; Wall Street Journal, 2008).
1.2.4 BUILDING SUPPORT FOR CLEAN
ENERGY POLICIES AND PROGRAMS
By quantifying and promoting the multiple benefits
of planned clean energy programs, states can address
barriers by raising awareness and building support
from key decision-makers and stakeholders by illumi-
nating strategic tradeoffs among energy resources. For
example, Connecticut's Climate Change Action Plan
is aimed at reducing GHG levels to 1990 levels by the
year 2010 and an additional 10% below that by 2020.
The plan evaluated 55 action items, including a large
number of clean energy activities. Connecticut found
that demonstrating the anticipated multiple benefits
early in the Action Plan development process, and
involving numerous stakeholders in this process, were
key to promoting the plan and obtaining the support of
multiple stakeholders (see text box Connecticut Incor-
porates Multiple Benefits in Evaluation Criteria for New
Capacity Additions) (CCC, 2005).
1.3 HOW DO STATES ASSESS THE
MULTIPLE BENEFITS OF CLEAN
ENERGY?
The preceding sections described how states are
advancing clean energy policies and programs and
the importance of assessing the multiple benefits of
these policies and programs. This section provides
an overview of how states conduct multiple benefits
analyses and key issues for states to consider as part of
the analyses.
Figure 1.3.1 illustrates the relationships among the
multiple benefits of clean energy. As shown in the
figure, while energy savings may be a primary goal of
clean energy policies and programs, other benefits also
accrue from these investments. These benefits are esti-
mated based, in part, on the energy savings estimates,
and in many cases may also be used as inputs for esti-
mating one or more of the other benefits.
It is not necessary for a state to evaluate all of the multi-
ple benefits of clean energy. Typically, a states priorities
and the purpose of its analysis influence which benefits
are of most interest. Understanding the relationship
between the benefits, however, can help states decide
how to go about evaluating the benefits of interest.
As an example of how the different benefits of clean
energy are related, consider a state that is contemplat-
ing a suite of energy efficiency programs. Based on
funding levels and assumptions about participation in
the programs, the state can estimate the direct energy
savings likely to accrue from them. The benefits, how-
ever, do not end there. A state can use the energy sav-
ings estimates to evaluate the benefits of the programs
on the state's energy system, economy, environment,
and public health. For example, the energy demand
reduction could be large enough to delay or eliminate
the need to construct new conventional power plants,
which can be quite costly. This would be a benefit to
the energy system. The decrease in the generation of
fossil-fuel-based electricity may result in a reduction in
GHG emissions and/or criteria air pollutants. Criteria
air pollutant reductions affect air quality and could
lead to public health benefits. These benefits can be
estimated and assigned an economic value. Consumers
would enjoy reduced energy costs, which could lead to
an increase in spending on non-energy products and
services. The economic benefits of the public health
improvements (e.g., improved productivity from re-
duced sick days), energy cost and system savings, and
investments in energy efficient equipment would likely
stimulate the economy and create jobs.
States can take the following steps when planning and
conducting an analysis of a clean energy policy, activ-
ity, or program that examines some or all of these clean
energy benefits:
Determine which clean energy goals, policies, activi-
ties, and/or programs to evaluate. When estimating
the multiple benefits of their clean energy policies
and programs, states can choose to focus on the
benefits of a single clean energy activity (e.g., ret-
rofitting a single state government building) or an
entire program (e.g., the state's portfolio of energy
efficiency activities, RPS, or green purchasing
program). The clean energy activities selected for
assessment can be identified, for example, based on
the state's overall energy policy and planning goals,
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 11
-------
FIGURE 1.3.1 RELATIONSHIP BETWEEN ENERGY SAVINGS & OTHER BENEFITS
OF CLEAN ENERGY INITIATIVES
ENERGY SYSTEM BENEFITS (Chapter 3)
Primary Electric System Benefits
Avoided generation.
Energy loss, and
System capacity
Secondary Electric System Benefits
Ancillary costs.
Reliability, and
Fuel diversification
I
AIR AND HEALTH BENEFITS (Chapter 4)
Criteria Air Pollutant
and/or Greenhouse Gas
DIRECT ENERGY Emission Benefits
IMPACTS (Chapter 2)
Pounds or tons of:
Change in kWh CO2,
supplied . PM,
Change in kWh . CO,
consumed <-Q
' N0x'
O ,
VOCs
Air Quality Benefits
Micrograms per cubic
meter ([ug/m3], or
Parts per million [ppm]
-ป *
J
Human Health Benefits
Changes in incidences of:
Mortality, bronchitis,
respiratory
Hospital admissions,
Upper and lower
respiratory symptoms.
and
Asthma effects
^
ECONOMIC BENEFITS (Chapter 5)
Direct Effects
Energy cost, waste heat or
displacement savings
Program Administrative, construction.
equipment, and operating costs
Sector transfers
Macroeconomic Benefits
Changes in:
Employment,
^^^^^ Gross state product.
Economic output.
Economic growth.
Personal income/earnings
^
regulatory or legislative requirements, or findings
from existing potential studies for energy efficiency
and/or renewable energy that provide important
information on which activities are most likely to
result in energy savings and other benefits.
1 Determine the goals and objectives of the multiple
benefits analysis. It is important to lay out the ra-
tionale for conducting a benefits analysis. Issues to
consider include:
> Why is the analysis being conducted? As de-
scribed in Section 1.4, there are many reasons
to analyze the benefits of a states clean energy
initiatives. For example, states can consider
whether the information will be used primar-
ily to gain support for their initiative; to help
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 12
-------
Modeling Approaches
This Resource describes a broad range of
modeling approaches that may be applied
to estimating energy savings, costs, emis-
sions and other impacts of clean energy
resources. In an effort to guide decision-
making, the Resource distinguishes be-
tween "sophisticated" modeling approach-
es that may require significant financial and
time commitments, and "basic" approaches
that require fewer resources and may more
easily be implemented by the state's own
staff. This distinction is somewhat impre-
cise, as model sophistication could actually
be judged along a very broad continuum;
nonetheless, the distinction helps convey in
broad strokes how approaches to multiple
benefit analyses can differ. For purposes of
this discussion:
Basic approaches (e.g., spreadsheet
analyses, trend extrapolations) tend to be
characterized by a relatively simple formu-
lation, such as the use of activity data (e.g.,
changes in generation levels) and factors
(e.g., emissions factors). In these approach-
es there is no attempt to represent the
underlying system (generation dispatch),
but instead they rely on factors or trends to
capture what would be expected to result.
In the example above, the emissions factor
is meant to represent the average of what
would actually be displaced by a clean en-
ergy resource that operates over a long pe-
riod of time and under varying conditions.
These factors and other inputs may be
based on the results of more sophisticated
modeling performed by others. Simpler
approaches can provide a reasonable level
of precision, depending on the nature and
source of the parameter. Each user will
have to assess whether the method and re-
sults are suitable for the intended purpose.
Sophisticated approaches tend to be char-
acterized by extensive underlying data and
relatively complex formulation that repre-
sents the fundamental engineering and eco-
nomic decision making of the entity (e.g.,
power sector system dispatch or capacity
expansion modeling), or complex physical
processes (such as in air dispersion model-
ing). Sophisticated models generally provide
greater detail than the basic methods, and
can capture the complex interactions within
the electricity market and with other mar-
kets or systems. They can be used to inform
discussions of what should happen (optimi-
zation) or what might happen given certain
assumptions (simulation). These approaches
are generally appropriate for short- or long-
term analyses, or analyses in which unique
demand and supply forecasts are needed to
incorporate the specific changes being con-
sidered (e.g., implementation of a renew-
able portfolio standard).
Regardless of what approach is chosen, it
is important to understand the strengths
and limitations of the method or model.
Specifically, it is important to recognize the
following:
Models are mathematical representations
of physical or economic processes in the
real world; therefore, these tools are only
as good as our understanding of these
processes. The results will be influenced
by the model formulation. For example, an
optimization model tells us what we should
do under the assumed conditions and rep-
resents the "best" or least cost approach.
A simulation model, potentially with logit
functions or market share algorithms, will
help us understand what might happen.
Simulation models offer insights into how a
complex system responds to changing con-
ditions and specific assumed conditions.
Data inputs and key driving assumptions
have a fundamental effect on the out-
comes, some more than others.
What actually occurs (or has occurred) will
depend on what values these key drivers
ultimately take. For all, there is some de-
gree of uncertainty: fuel prices, weather,
unit availability, load levels and patterns,
technology performance, future market
structure and regulatory requirements,
to name only a few, all have considerable
uncertainties surrounding them. However,
the strength of models, particularly those
bottom-up models with engineering-
economic detail, is that they provide a con-
sistent framework for understanding how a
system responds to different stimuli and to
characterize the uncertainty surrounding
our best estimates.
design a clean energy program and select the
specific activities to include in the program,
provide data for a regulatory purpose (e.g., a
SIP or cap and trade program); or to support
related environmental, planning, or economic
development policy and program decisions.
Which benefits will be analyzed? States can
concentrate on estimating some or all of the
multiple benefits of their clean energy activ-
ity or program, depending on the purpose
and scope of the initiative. This decision will
depend on the audience and their interests,
available financial and staff resources, and the
type and scope of the clean energy initiative(s)
being assessed. For example, when decid-
ing whether to conduct an energy efficiency
retrofit of a single building, states may want
to estimate the energy savings and GHG
emission reductions of other building retrofit
options and use this information to select
the likely candidate for retrofitting. When
developing a clean energy plan or assessing a
more extensive clean energy initiative, it may
be more appropriate to assess a broad range of
benefits and use this information to help build
widespread support for the program.
Determine how to conduct the analysis. Multiple
benefits analyses can employ a variety of ap-
proaches, ranging from basic screening estimates
and spreadsheet analyses to more sophisticated
modeling approaches. States will consider a variety
of issues when determining the most appropriate
approach for their needs and circumstances, and
will balance competing factors as necessaryfor
example, the scope and rigor of the analysis may
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 13
-------
be balanced against the level of resources available.
Key issues include:
> What financial and staff resources are available?
> What other kinds of expertise (e.g., in-house
staff and outside consultants) are available?
Do data exist from similar analyses or for other
states or regions? Or will a new analysis be
required?
Is the analysis retrospective (an historical assess-
ment) or prospective (forward-looking)?
What level of rigor is required? Is it for regula-
tory purposes or a preliminary screening of
options?
> Will the analysis entail an iterative approach
where the state explores a wide range of options
using screening methods and then conducts a
more comprehensive analysis of only the most
promising options?
More detailed information about how to estimate
the potential benefits of clean energy initiatives is
presented in the remaining chapters of the Resource, as
follows:
Chapter 2: Assessing the Potential Energy Impacts
of Clean Energy Initiatives.
Chapter 3: Assessing the Electric System Benefits of
Clean Energy Initiatives.
Chapter 4: Assessing the Air Pollution, Greenhouse
Gas, Air Quality, and Health Benefits of Clean
Energy Initiatives.
Chapter 5: Assessing the Economic Benefits of
Clean Energy Initiatives.
Each chapter describes approaches for calculating or
estimating prospective benefits based on varying levels of
rigor and provides examples of states' experiences using
multiple benefits analysis to promote clean energy. The
chapters provide general information on how to conduct
and evaluate analyses of multiple benefits, rather than
serving as a detailed workbook for quantifying benefits.
Taken as a whole, these chapters provide a framework
for states to use in determining the likely benefits of their
clean energy goals, policies, and programs and using this
information to support these initiatives.
References URL Address
Bezdek, Roger. 2007. Renewable Energy and Energy Efficiency: Economic
Drivers for the 21st Century. Prepared for the American Solar Energy Society.
Bluestein, J., E. Salerno, L. Bird, and L. Vimmerstedt. 2006. Incorporating Wind
Generation in Cap and Trade Programs. National Renewable Energy Laboratory
(NREL) Technical Report. NREL/TP-500-40006. July.
Burtraw, D., A. Krupnick, K. Palmer, A. Paul, M. Toman, and C. Bloyd. 2001.
Ancillary Benefits of Reduced Air Pollution in the United States from Moderate
Greenhouse Gas Mitigation Policies in the Electricity Sector. Resources for the
Future, Discussion Paper 01-61. December.
Connecticut Climate Change. 2005 Connecticut Climate Action Plan Web Site.
CPUC. 2003. Energy Action Plan. Adopted May 8.
Cinnamon, B., T. Beach, M. Huskins, and M. McClintock. 2005. The Economics
of Solar Power for California. White Paper. August.
Clean Air Report (via lnsideEPA.com). 2006. States, Utilities Divided Over Role
of Carbon In Building New Plants. CLEANAIR-17-25-10. December 14.
http://www.greenforall.org/resources/renewable-
energy-an d-en ergy- efficien cy- economic
http://www.eere.energy.gov/windandhydro/
windpoweringamerica/pdfs/wpa/policy_cap_trade.pdf
h ttp://www.rff. org/Documen ts/RFF- DP-01-61.pdf
http://www.ctclimatechange.com/StateActionPlan.html
http://docs.cpuc.ca.gov/published/report/28715.htm
http://akeena.net/Library/pdf s/Economics_of_Solar_
Power_for_California_Aug2005.pdf
N/A
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 14
-------
1
References
Clemmer, S., B. Grace, and K. Cory. 2003. A Study to Evaluate the Impacts
of Increasing Wisconsin's Renewable Portfolio Standard. Prepared for the
University of Wisconsin-Madison and the Wisconsin Division of Energy.
October.
Conaway, C. 2006. The Challenge of Energy Policy in New England. Federal
Reserve Bank of Boston, New England Public Policy Center. April.
Deyette, J., S Clemmer. 2005. Increasing the Texas Renewable Energy
Standard: Economic and Employment Benefits. Prepared for Union Of
Concerned Scientists.
DRAM. 2005. Demand Response and Advanced Metering Coalition.
EIA. 2008. Annual Energy Outlook 2008. DOE/EIA-0383(2008). U.S. Energy
Information Administration, Washington, DC. June.
Energy Trust of Oregon. 2005. Cost- Effectiveness Policy and General
Methodology for the Energy Trust of Oregon (4.06.000-P).
Freese, B. and S. Clemmer. 2006. Gambling with Coal: How Future Climate
Laws Will Make New Coal Power Plants More Expensive. Union of Concerned
Scientists. September.
Georgia Environmental Facilities Authority (GEFA). 2006a. Georgia Energy
Review 2005. Prepared by the GEFA Division of Energy Resources. March.
Georgia Environmental Facilities Authority (GEFA). 2006b. State Energy
Strategy for Georgia. Prepared by the GEFA Division of Energy Resources.
September 1.
Georgia Office of Communications. 2006. Press Release: Governor Announces
Development of Statewide Energy Strategy. March 2.
Grover, S. 2005. Economic Impacts of Oregon Energy Tax Credit Programs
(BETC/RETC). Prepared by ECONorthwest for the Oregon Department of
Energy. February.
Haberl, J., C. Culp, B. Yazdani, D. Gilman, T. Fitzpatrick, S. Muns, M. Verdict, M.
Ahmed, Z. Liu, J. Baltazer-Cervantes, J. Bryant, L. Degelman, D. Turner. 2004.
Energy Efficiency/Renewable Energy Impact in the Texas Emissions Reduction
Plan (TERP). ESL-TR-04/12-01. December.
Jacobson, D., P. O'Connor, C. High, and J. Brown. 2006. Final Report on
the Clean Energy/Air Quality Integration Initiative Pilot Project of the U.S.
Department of Energy's Mid-Atlantic Regional Office. DOE/GO-102006-2354.
August.
Jensen, V., and E. Lounsbury. 2005a. Assessment of Energy Efficiency Potential
in Georgia. Prepared for the Georgia Environmental Facilities Authority by ICF
Consulting. May.
Jensen, V., and E. Lounsbury. 2005b. Strategies for Capturing Georgia's
Energy Efficiency Potential. Prepared for the Georgia Environmental Facilities
Authority by ICF Consulting. May.
Kammen, D., K. Kapadia, M. Fripp. 2004. Putting Renewables to Work: How
Many Jobs Can the Clean Energy Industry Generate? April.
1
h ttp://energytaskforce. wi.gov/docview.asp ?docid=2
http://www.bos.frb.org/economic/neppc/wp/2005/
neppcwp0502.htm
http://www.ucsusa.org/assets/documents/clean_
energy/texas_res_report-02-05_final.pdf
http://www.dramcoalition.org/idl9.htm
http://www.eia.doe.gov/oiaf/archive/aeo08/index.html
http://energytrust.0rg/library/policies/4.06.000.pdf
http://www.ucsusa.org/assets/documents/clean_
energy/gambling_with_coal_final_report_sept_06.pdf
http://www.gefa.org/lndex.aspx?page=93
http://www.gefa.org/lndex.aspx?page=93
http://gov.georgia.gov/00/press/detail/0,2668,
78006749_90418617_91273717,00.html
http://www.oregon.gov/ENERGY/CONS/docs/EcoNW_
Study.pdf
h ttp://esi eslwin.tam u. edu/docs/documen ts/tceq-
report-2-14-2005-vol-l.pdf
http://www.eere.energy.gov/wip/pdfs/40477.pdf
http://www.gefa.org/Modules/ShowDocument.aspx?
documentid=46
http://www.gefa.org/Modules/ShowDocument.aspx?
documentid=45
http://rael.berkeley.edu/old-site/renewables.jobs.2006.
pdf
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 15
-------
1
References
Nadel, S., A. Shipley, and R.N. Elliot. 2004. The Technical, Economic and
Achievable Potential for Energy- Efficiency in the U.S. - A Meta-Analysis
of Recent Studies. American Council for an Energy- Efficient Economy,
Washington, DC. From the proceedings of the 2004 ACEEE Summer Study on
Energy Efficiency in Buildings.
New England Governors and Eastern Canadian Premiers (NEG-ECP). 2003.
Resolution 28-7: Resolution Concerning Environmental Projects and Issues.
September.
New England Governors and Eastern Canadian Premiers (NEG-ECP). 2006.
Resolution 30-2: Resolution Concerning Energy. May.
New England Public Policy Center (NEPPC) and Federal Reserve Bank of
Boston. 2005. Fueling the Future: Energy Policy in New England Conference
Report 05-2. December.
NYSERDA. 2006. New York Energy Smart5" Program Evaluation and Status
Report Year Ending December 31, 2005. Report to the System Benefits Charge
Advisory Group. Final Report. May.
NYSERDA. 2008. New York Energy $martSM Program Evaluation and Status
Report: Year Ending December 31, 2007. Report to the System Benefits Charge
Advisory Group. Final Report. March.
Northeast International Committee on Energy (NICE). 2005. 2004-2005 Report
to the 29th Conference of New England Governors and Eastern Canadian
Premiers on Energy. August.
Northeast International Committee on Energy (NICE). 2006. Report to the New
England Governors and Eastern Canadian Premiers on the Activities of the
Northeast Committee on Energy. May.
Pennsylvania. 2008. Office of the Governor Web Site. Gov. Rendell Signs
Bill Establishing $650 Million Energy Fund to Support Conservation, Spur
Renewable Energy Development. July 9.
Prindle, W., A. Shipley, and R. Elliott. 2006. Energy Efficiency's Role in a Carbon
Cap-and-Trade System: Modeling the Results from the Regional Greenhouse
Gas Initiative. ACEEE Report E064. May.
PSE IRP. 2007. Puget Sound Energy 2007 Integrated Resource Plan. Filed with
the Washington Utilities and Transportation Commission May 31.
Rabe, B.C. 2002. Greenhouse & Statehouse: The Evolving State Government
Role in Climate Change. Pew Center on Global Climate Change. November.
Regulatory Assistance Project (RAP). 2005. Electric Energy Efficiency and
Renewable Energy in New England: An Assessment of Existing Policies and
Prospects for the Future. Prepared by The Regulatory Assistance Project and
Synapse Energy Economics, Inc. May.
Singh, V. and J. Fehrs. 2001. The Work That Goes Into Renewable Energy.
Renewable Energy Policy Project Research Report. November.
State and Territorial Air Pollution Program Administrators/Association of
Local Air Pollution Control Officials (STAPPA/ALAPCO). 2005. Alternative NOX
Allowance Allocation Language for the Clean Air Interstate Rule. Prepared by
the State and Territorial Air Pollution Program Administrators (STAPPA) and the
Association of Local Air Pollution Control Officials (ALAPCO). August.
1
URL Address
http://www.aceee.org/conf/04ss/rnemeta.pdf
http://www.negc. org/03resolutions/res28_7.html
http://www.negc. org/resolutions/Res_ 30-2_5-06.pdf
http://www.bos.frb.org/economic/neppc/
conreports/2005/conreport052.pdf
h ttp://www.nyserda. org/Energy_
Information/OBsbcreport.asp
http://www.nyserda.org/pdfs/Combined%20Report.pdf
http://www.cap-cpma.ca/images/pdf/eng/2005 NICE
Report Finalpdf
http://www.cap-cpma.ca/images/pdf/eng/NICE 2006
Annual Report.pdf
http://www.depweb.state.pa.us/news/cwp/view.
asp?a=3&q=538660
http://aceee.org/pubs/e064.pdf?CFID=978956&CFTO
KEN=15830524
http://www.pse.com/ENERGYENVIRONMENT/
ENERGYSUPPLY/Pages/pse2007irpView.aspx
h ttp://www.pewclima te. org/docilploads/sta tes_
greenhouse.pdf
http://www.raponline.org/Pubs/RSWS-EEandREinNE.pdf
http://www.repp.org/articles/static/l/binaries/LABOR_
FINAL_REV.pdf
h ttp://www.4cleanair. org/ Blues tein-cairalloca tion -
final.pdf
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 16
-------
1 1
References URL Address
Stoddard, L, J. Abiecunas, and R. O'Connell. 2006. Economic, Energy, and
Environmental Benefits of Concentrating Solar Power in California. Prepared
by Black & Veatch for U.S. DOE National Renewable Energy Laboratory. April.
Sumi, D., G. Weisbrod, B. Ward, and M. Goldberg. 2003. An Approach to
Quantifying Economic and Environmental Benefits for Wisconsin's Focus on
Energy. Presented at International Energy Program Evaluation Conference. August.
SWEEP. 2002. The New Mother Lode: The Potential for More Efficient
Electricity Use in the Southwest. Southwest Energy Efficiency Project, Report
for the Hewlett Foundation Energy Series. November.
http://www.nrel.gov/docs/fy06osti/39291.pdf
h ttp://www.edrgroup. com/pdf/sumi- weisbrod- wis-
energy-iepec.pdf
http://www.swenergy.org/nml
U.S. Environmental Protection Agency. 1999. Guidance on Establishing an http://www.epa.gov/cleanenergy/energy-programs/
Energy Efficiency and Renewable Energy (EE/RE) Set-Aside in the NOX Budget state-and-local/guidance.html
Trading Program. EPA Office of Atmospheric Programs and Office of Air and
Radiation. March.
U.S. Environmental Protection Agency. 2004a. Guidance on State
Implementation Plan (SIP) Credits for Emission Reductions from Electric-
Sector Energy Efficiency and Renewable Energy Measures. EPA Office of Air
and Radiation. August.
U.S. Environmental Protection Agency. 2004b. Incorporating Emerging and
Voluntary Measures in a State Implementation Plan. EPA Office of Air and
Radiation. September.
U.S. Environmental Protection Agency. 2004c. Output-Based Regulations: A
Handbook for Air Regulators. EPA Office of Atmospheric Programs, Climate
Protection Partnerships Division. August.
U.S. Environmental Protection Agency. 2005a. Incorporating Bundled Emission
Reduction Measures in a State Implementation Plan. EPA Office of Air and
Radiation. August.
U.S. Environmental Protection Agency. 2005b. State Set-Aside Programs for
Energy Efficiency and Renewable Energy Projects Under the NOX Budget
Trading Program: A Review of Programs in Indiana, Maryland, Massachusetts,
Missouri, New Jersey, New York, and Ohio (Draft Report). EPA Office of
Atmospheric Programs. September.
U.S. Environmental Protection Agency. 2006. Clean Energy-Environment
Guide to Action: Policies, Best Practices, and Action Steps for States. April.
U.S. Environmental Protection Agency. 2007. Emissions & Generation Resource
Integrated Database (eGRID), eGRID2007 Version 1.1.
U.S. Environmental Protection Agency. 2008. Clean Energy Policy Maps Web
Site. August.
Vermont. 2005. Vermont State Agency Energy Plan for State Government.
Wall Street Journal. 2008. Market Watch. PA Governor Rendell Says Clean
Energy Projects Will Create Jobs, Promote Conservation, Efficiency. October 20.
Weisbrod, G. and B. Weisbrod. 1997. Measuring Economic Impacts of Projects
and Programs. Economic Development Research Group. April.
http://www.epa.gov/ttn/oarpg/tl/memoranda/
ereseerem_gd.pdf
http://www.epa.gov/ttncaaal/tl/memoranda/evm_
ievm_g.pdf
http://www.epa.gov/chp/documents/obr_final_9105.
pdf
http://www.epa.gov/ttn/oarpg/tl/
memoranda/10885guideibminsip.pdf
http://www.epa.gov/cleanenergy/documents/eere_rpt
pdf
http://www.epa.gov/statelocalclimate/resources/
action-guide.html
http://www.epa.gov/egrid
http://www.epa.gov/statelocalclimate/state/tracking/
index.html
http://bgs.vermont.gov/sites/bgs/files/pdfs/BGS-
VTStateEnergyPlan.pdf
http://www.marketwatch.com/news/story/
pa-governor-rendell-says-clean/story.
aspx?guid={49A81C90-F822-4FBC-88F5-
55B6DOD8899E}6fdist=hppr
http://www.smartbrief.com/news/aaaa/industryPR-
detail.jsp?id=lC2429AC-1174-48EO-8FB3-
E84476F6C1C3
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 17
-------
URL Address
Weisbrod, G., K. Polenske, T. Lynch, and X. Lin. 1995. The Long-Term Economic
Impact of Energy Efficiency Programs and Renewable Power for Iowa: Final
Report. Economic Development Research Group, Boston, MA. December.
http://www.edrgroup.com/library/energy-
environment/iowa-energy.html
Western Governors'Association (WGA). 2006a. Energy Efficiency Task Force
Report. Final report to the Clean and Diversified Energy Advisory Committee
(CDEAC). January.
h ttp://www. westgov. org/wga/initia tives/cdeac/Energy
Efficien cy- full.pdf
Western Governors'Association (WGA). 2006b. Clean Energy, a Strong
Economy and a Health Environment. Report of the Clean and Diversified
Energy Advisory Committee (CDEAC) to the Western Governors. June.
http://www.westgov.org/wga/meetings/am2006/
CDEAC06.pdf
Wisconsin Office of the Governor. 2006. Press Release: Governor Doyle Signs
Senate Bill 459, Bipartisan Energy Efficiency and Renewables Legislation.
March 17.
http://www.wisgov.state.wi. us/journal_media_detail.
asp?locid = 196-prid = 1830
Wisconsin Public Service Commission. 2007. Public Service Commission of
Wisconsin Focus on Energy Evaluation Semiannual Report (FY07, Year-end)
Final: September 11, 2007 Revised: November 1.
http://www.focusonenergy.com/files/
Documen t_Managemen t_ System/Evalua tion/
semiannualsecondhalffy07_evaluationreport.pdf
Wiser, R., M. Bolinger, and M. Clair. 2005. Easing the Natural Gas Crisis:
Reducing Natural Gas Prices through Increased Deployment of Renewable
Energy and Energy Efficiency. Ernest Orlando Lawrence Berkeley National
Laboratory (LBNL). January.
http://eetd.lbl.gov/ea/ems/reports/56756.pdf
Xcel Energy. 2006. Petition to the Minnesota Public Utilities Commission to
Initiate a Competitive Resource Acquisition Process for 375 MW of Base Load
Generation. November.
h ttps://www. edocke ts.state.mn. us/EFiling/Sh o wFile.
do?DocNumber=4717612
CHAPTER 1 | Assessing the Multiple Benefits of Clean Energy 18
-------
CHAPTER TWO
Assessing the Potential Energy
Impacts of Clean Energy Initiatives
Amidst rising concerns about energy prices, the avail-
ability of reliable energy resources, air quality, and cli-
mate change, many states across the country are using
clean energy policies to help meet their expanding elec-
tricity demand in a clean, low-cost, reliable manner.
Nearly 40 states are using planning and incentive
structures to promote clean energy within their
own operations;
More than 30 states have adopted a number of regu-
latory and market-based energy efficiency actions
that increase investment in cost-effective energy
efficiency by consumers, businesses, utilities, and
public agencies; and
More than 40 states have taken energy supply ac-
tions to support and encourage continued growth
of clean energy supply1
These actions result in measurable reductions in de-
mand for conventional fossil-fuel-powered electricity
as well as reductions in natural gas used for heating,
and/or an increase in the amount of electricity gener-
ated with clean, renewable energy sources.
This chapter provides state policymakers with methods
and examples they can use to estimate the potential di-
rect energy impacts of electricity-related clean energy
options for policy and program planning purposes. By
understanding the potential energy savings of these
programs and policies, state officials can:
Demonstrate the energy-related impacts of existing
and potential clean energy programs;
1 For more information about which states have implemented these policies,
see: http://www.epa.gov/statelocalclimate/state/tracking/mdex.html
U
o
Q
CHAPTER ONE
Introduction
CHAPTER TWO
Potential Energy Impacts of Clean Energy
CHAPTER THREE
Electric System Benefits of Clean Energy
CHAPTER FOUR
Air Quality Benefits of Clean Energy
CHAPTER FIVE
Economic Benefits of Clean Energy
APPENDIX A
Catalogue of Clean Energy Case Studies
APPENDIX B
Tools and Models Referenced in Each Chapter
CHAPTER TWO CONTENTS
2.1 How Do Clean Energy Policies Affect Energy? 20
2.2 How Can States Estimate the Potential Direct Energy
Impacts of Clean Energy Policies? 21
2.2.1 STEP 1: Develop a Business-As-Usual
Energy Forecast 22
2.2.2 STEP 2: Quantify Implications of Targets
and Goals 33
2.2.3 STEP 3: Estimate Potential Direct Energy
Impacts 35
2.2.4 STEP 4: Create an Alternative Policy Forecast 44
2.3 Case Studies 45
2.3.1 Texas Building Code 45
2.3.2 Vermont - Energy and Energy Savings
Forecasting 47
Information Resources 49
References 50
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 19
-------
STATES ARE QUANTIFYING POTENTIAL DIRECT ENERGY
IMPACTS OF CLEAN ENERGY INITIATIVES
The New York Energy Smart5" public benefits program, funded
through a systems benefit charge, was implemented in 1998
to improve New York's energy reliability, reduce energy costs,
mitigate environmental and public health effects related to
energy use in New York, and enhance the state economy
(NYSERDA, 2008). Each year, the New York State Energy
Research and Development Authority (NYSERDA) develops a
report for the New York State Public Service Commission on
the energy savings and progress toward program and energy
savings goals.
Between 1998 and 2004, the program achieved cumulative:
electricity savings of 1,400 GWh, and
energy cost savings of $195 million.
The program was extended in 2005 for an additional five years
and the annual budget increased from $150 million to $175
million (NYSERDA, 2005; NYSERDA, 2008). The expanded
program continues to achieve significant benefits. By year-end
2007, the overall program had achieved more than 3,000 GWh
of electricity savings.
Based on these electricity savings estimates and related
investments, NYSERDA calculated the cumulative benefits of
the Energy $martSM program through 2007 and found that it:
Reduced annual energy bills by $570 million for
participating customers,
Created and retained 4,700 jobs,
Reduced nearly 2,600 and 4,700 tons of NOX and SO2
respectively, and
Decreased annual CO2 emissions by 2 million tons
(NYSERDA, 2008).
Using projections of New York's clean energy investments
and electricity savings, NYSERDA estimated that by 2027 the
program will create more than 7,200 jobs, increase labor
income more than $300 million each year, and increase total
annual output in the state by $503 million. This information
about progress and benefits will inform future decisions about
New York Energy $martSM program funding (NYSERDA, 2008).
Evaluate the implications of new goals, targets, or
legislative actions;
Evaluate the feasibility of or progress toward clean
energy-related goals or standards;
Evaluate the actual and potential effectiveness of
technology- or sector-specific clean energy pro-
grams in achieving energy savings;
Compare across clean energy options; and
Evaluate the actual and potential co-benefits of
clean energy policies, including benefits to the
energy system, economy, environment, and public
health.
As illustrated in the text box States are Quantifying
Potential Direct Energy Impacts of Clean Energy Initia-
tives, estimates of potential energy savings serve as a
foundation for subsequent analysis of multiple benefits
and help demonstrate the value of a program. States
can conduct similar analyses of their clean energy pro-
grams using methods and tools described in the rest of
this chapter.
Section 2.1 provides a brief explanation of how
clean energy initiatives affect energy use and elec-
tricity generation requirements.
Section 2.2 describes methods for estimating
potential energy savings or renewable energy gen-
eration. This section and the remaining chapters
of the Resource focus on prospective, rather than
retrospective, analyses. See the text box Retrospec-
tive versus Prospective Calculation of Energy Savings
for more information.
Section 2.3 presents case studies that illustrate how
states have used some of these approaches to devel-
op a baseline, a business as usual (BAU) forecast,
and energy savings or renewable energy forecasts
while planning their clean energy policies.
2.1 HOW DO CLEAN ENERGY
POLICIES AFFECT ENERGY?
The two primary objectives of clean energy initiatives
are typically to:
1. Implement low-cost energy efficiency measures
that reduce the demand for energy, and/or
2. Deploy renewable energy systems (both thermal
and electric) or highly efficient cogeneration
systems to meet energy demand with the cleanest
resources available.
Energy efficiency initiatives include energy efficiency
savings goals; energy efficiency portfolio standards;
public benefit funds for energy efficiency; building
codes; appliance standards; revolving loan programs for
energy efficiency; energy performance contracting; and
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 20
-------
FIGURE 2.2.1 STEPS TO ESTIMATING ENERGY
IMPACTS OF CLEAN ENERGY
STEPl
Develop a BAU Energy Forecast
STEP 2
Quantify Implications of Targets and Goals
STEPS
Estimate Potential Direct Energy Impacts
STEP 4
Create an Alternative Policy Forecast
incentives, grants and rebates for efficiency.2 Through
regulatory, market-based, and voluntary approaches,
these programs are designed to advance the deploy-
ment of energy efficient technologies. The outcome
of efficiency efforts is measured in terms of reduced
end-use consumption or energy savings (in kWhs or
Btus) and peak demand (MW, or maximum Btu/hour),
which reduce the amount of energy demanded from
generators or delivered from natural gas producers.3
Renewable energy initiatives include renewable elec-
tricity generation and energy goals; renewable energy
portfolio standards; public benefit funds for renewable
energy; and revolving loan programs, incentives, and
grants and rebates for renewable energy investments.
Through regulatory, market-based, and voluntary ap-
proaches, these programs are designed to advance the
deployment of renewable energy fuels and technolo-
gies. Power produced by renewable energy generators
displaces supply from existing or planned fossil-fueled
electricity generation, sometimes described as "avoid-
ed energy."4
2 These and other clean energy activities are described in the EPA Clean
Energy-Environment Guide to Action: Policies, Best Practices, and Action
Steps for States (U.S. EPA, 2006).
3 As noted in Chapter 1, while clean energy resources include energy ef-
ficiency and energy resources that reduces demand for electricity and fossil
fuels, the focus of this Resource is on those that affect electricity demand and/
or the electric system.
4 The actual impact of incremental renewable energy production on the
energy system as a whole is complex and depends on factors such as the timing
of production and the baseload requirements of the power grid. These energy
system impacts are discussed in Chapter 3.
RETROSPECTIVE VERSUS PROSPECTIVE CALCULATION OF
ENERGY SAVINGS
States can assess energy impacts from two perspectives:
retrospectively, to evaluate impacts of existing investments, or
prospectively, to plan new or modified initiatives. This Resource
describes prospective techniques for estimating energy savings
or renewable energy generation to help states plan: that is,
methods and models that calculate energy impacts expected
to occur in the future as a result of the state's proposed clean
energy initiatives. Prospective analyses of energy impacts are
appropriate, for example, when a state wants to gain support
for a proposed clean energy policy, is assessing the relative
costs and benefits of alternative policies in order to select the
most cost-effective clean energy approach, or is determining
the budget level required to meet clean energy goals.
A retrospective approach, in contrast, is based on
measurements of actual impacts that have already accrued
from the state's clean energy actions. Actual energy savings
from energy efficiency programs, for example, are calculated
using "measurement and verification" (M&V) methods, whereby
measurements determine actual savings from measures
implemented within an individual facility. Energy savings are
calculated using the following approach:
Select a representative sample of projects.
Determine the savings of each project in the sample,
based on deemed savings values (i.e., claimed savings) or
measured savings, energy bills, or calibrated computer
simulation.
Apply the sample project's savings to the entire population
(e.g., the clean energy program).
More information about retrospective calculation of energy
savings from energy efficiency is available in the National Action
Plan for Energy Efficiency (NAPEE), Model Energy Efficiency
Program Impact Evaluation Guide, November 2007 (http://
www.epa.gov/cleanenergy/documents/evaluation_guide.pdf)
and U.S. EPA Lead by Example Guide, June 2009 (http://www.
epa.gov/statelocalclimate/resources/example.html).
These direct energy supply impact estimates are the
foundation for calculating potential cost savings and
other benefits to the state economy, energy system ben-
efits, and environmental and public health benefits.
2.2 HOW CAN STATES ESTIMATE THE
POTENTIAL DIRECT ENERGY IMPACTS
OF CLEAN ENERGY POLICIES?
There are four primary steps for estimating the poten-
tial direct energy impacts from clean energy policies
(see Figure 2.2.1). The first step is to establish a BAU
forecast of energy supply and demand. This involves
taking a look at the historical demand and supply
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 21
-------
portfolio within a state (i.e., developing the baseline)
and projecting it forward, based on assumptions about
the future. The projection is a BAU forecast that il-
lustrates what state energy demand, consumption, and
supply will most likely be in the absence of additional
clean energy policies beyond those already considered
in resource planning.
This projection can be used in a second step to develop
or quantify the implications of an energy-related target
if a state is interested.
The third step is to estimate the energy savings (or
clean energy supply) from a proposed clean energy
initiative or portfolio of initiatives. The energy savings
are determined by estimating the impact on energy
consumption levels and patterns of a specific policy ap-
proach, or the energy output from renewable resources.
The fourth step, creating an alternative policy forecast,
allows the state to consider potential outcomes of real-
izing the direct energy impacts. In the case of efficien-
cy, the energy savings estimates are subtracted from the
BAU forecast developed under Step 1 to create a new
energy forecast. For clean energy supply alternatives,
the impacts estimates are used to assess impacts on the
electric power system (in terms of what is displaced
that otherwise would have been operated).
Because there are so many details and assumptions
involved in estimating savings and creating alternative
policies, a state must choose the right approach for the
decision process at hand. As described below, the level
of available resources (including budget, personnel, and
data) often guides which approach to select when devel-
oping an energy savings estimate. For a quick compari-
son of policy alternatives, a top-down approach maybe
acceptable, while a bottom-up approach may be more
appropriate for program planning and budget setting.
Each step is described below.
2.2.1 STEP 1: DEVELOP A BUSINESS-AS-
USUAL ENERGY FORECAST
An energy baseline and BAU forecast documents the
historical, current, and projected pattern of energy sup-
ply and demand within a state. The BAU forecast illus-
trates what state energy use will look like in the future,
in the absence of additional policies beyond those al-
ready in place and planned. It typically includes current
programs, such as regulations, standards, or energy ef-
ficiency programs. The BAU forecast is a reference case
against which to measure the energy impacts of policy
initiatives or unexpected system shocks (e.g., severe
weather-related disruptions in energy supply).
As presented in Figure 2.2.2, the following six broad
steps are involved in developing a BAU energy forecast:
1. Define objectives and parameters;
2. Develop a historical energy baseline;
3. Choose method to develop the forecast or project
the historical energy baseline into the future;
4. Determine assumptions and review data;
5. Apply the chosen model or approach; and
6. Evaluate forecast output.
These six steps are described below.
STEP 1.1: Define Objectives and Parameters
For this chapter's purposes, the objective of the BAU
forecast is to aid in determining energy savings from
clean energy initiatives by offering a current and pro-
jected energy picture. To this end, states should:
Determine if the forecast will be short- or long-
term, and end-use based or sector-wide (i.e.,
explicitly modeling the building stock and end-use
equipment vs. using a top-down model of the total
sectoral or economy-wide demand);
Establish the level of rigor necessary;
Consider the availability of financial, labor, and
time resources to complete the forecast; and
Verify the amount of energy data readily available
to develop the forecast.
These factors will help states choose between basic and
more sophisticated forecasting approaches.
STEP 1.2: Develop a Historical Baseline
A comprehensive energy baseline includes the follow-
ing historical energy data:
Consumption (demand) by sector or fuel, and
Energy generation (supply) by fuel and/or
technology.
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 22
-------
Consumption data are often broken down by the sec-
tors that consume the fuels, including the commercial,
residential, industrial, transportation, and utility
sectors. This type of top-down baseline helps a state
understand the large and small consumers within a
state and helps target sectors for policy interventions.
Each sector can also be further disaggregated to show
the types of consumption within.
A top-down approach would be appropriate if a state
plans to evaluate or quantify the requirements of a
broad, state-wide energy efficiency or renewable en-
ergy goal. For example, in 2006, Wisconsin Governor
Jim Doyle launched the Declaration of Energy Inde-
pendence, which included a goal of using renewable
energy to generate 25 percent of the state's electricity
and 25 percent of its transportation fuels by 2025
(the "25x25" goal). Figure 2.2.3 illustrates a demand
baseline by sector that the Wisconsin Office of Energy
Independence developed to help it understand impli-
cations for energy consumption as it strives to achieve
its goal. This top-down baseline helped the state
understand how its total energy consumption (i.e.,
electricity, natural gas, petroleum, coal, and renewable
energy use) is spread across sectors and identify
which sectors seem most appropriate for further
investigation and potential program intervention
(Wisconsin, 2007).
An alternative or a complement to the top-down ap-
proach is to develop a bottom-up baseline. A bottom-
up baseline is very data-intensive, but provides more
information about activities within a particular sector
than an aggregated, top-down baseline that is used to
reveal trends and opportunities across sectors.
The bottom-up approach is most appropriate if a state
is exploring a sector- or technology-specific clean
energy policy. For example, if Wisconsin targets the
residential sector to help achieve its 25x25 goal, the
state could develop a bottom-up baseline that depicts
the amount of residential consumption attributed
to hot water heating, appliances, and cooling. If it
finds that the majority of residential consumption is
related to specific end-use equipment, it might focus
its program design efforts on the most cost-effective
and efficacious opportunities for equipment within the
residential sector.
It is important to recognize that both past and future
demand for energy are products of the economic and
weather conditions of the state as well as the types and
F
I
IGURE 2.2.2 SAMPLE FRAMEWORK FOR
)EVELOPING AN ENERGY FORECAST
1. Define Objectives
and Parameters
/
/
ฃ
t 2. Develop Historical
' Energy Baseline
' 2
/ jT 3. Choose N.
/ >/ Forecast >*
' r\. Method }
1 N. S=Supply /
1 Basic N. D= Demand / Sophisticated
1 Methods \ / Methods
' (AllSorD) N. /
1 r i '
Compile forecasts (SD)
Adopt Forecasts
Nominal Group
Technique
\
\ Linear/Nonlinear
\ Extrapolation
I
Time Series (D)
End Use fD}
Econometric (D)
Dispatch (S)
Capacity
Expansion (S)
\ 4. Determine
* XV Review Data
\ ' 1
\ 1 T
^ ' 5. Apply Model or
x ' Approach
\ *
N \
X^ \ ^
6. Evaluate
Forecast
Output
^ '
Demand
and/or Supply
Forecast
efficiencies of end-use appliances and equipment. Thus,
future forecasts often need a specific economic projec-
tion as a starting point and should assume normal
weather conditions.
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 23
-------
FIGURE 2.2.3 WISCONSIN RESOURCE ENERGY
CONSUMPTION BY ECONOMIC SECTOR
(Trillions of Btu and Percent of Total)
Transportation
437 (25%)
Agricultural
37 (2%)
Source: 2007 Wisconsin Energy Statistics, Wisconsin Office of Energy
Independence, Achieving 25x25, page 8.
FIGURE 2.2.4 WISCONSIN ELECTRIC UTILITY
GENERATING CAPACITY BY TYPE OF PLANT, 2005
(MW and Percent of Total)
Source: 2007 Wisconsin Energy Statistics, Wisconsin Office of Energy
Independence, Achieving 25x25, page 52.
FIGURE 2.2.5 WISCONSIN RESOURCE ENERGY
CONSUMPTION, BY TYPE OF FUEL, 2006
(Trillions of Btu and Percent of Total)
Electric Imports
124(7.1%)
"Electric imports" is the estimated resource energy used in other states or Canada to
produce the electricity imported into Wisconsin.
Source: 2007 Wisconsin Energy Statistics, Wisconsin Office of Energy
Independence, Achieving 25x25, page 6.
On the supply side, electricity generation data can also
be categorized by fuel type and sector.5 Figure 2.2.4
illustrates Wisconsin's supply side baseline that shows
electricity generation by type of fuel for a single year. A
baseline energy forecast requires data about the types
and amounts of fuel used to generate electricity, includ-
ing uranium; coal; natural gas; municipal solid waste;
wood; landfill gas; hydro; and petroleum fuels, such as
distillates and residuals. Depending on a state's defini-
tion of "renewable," renewable fuels can include wood,
landfill gas, pyrolysis liquid/gas, geothermal, hydro,
solar PV/thermal, wind, and municipal solid waste.
Electricity generation data typically include electricity
generation that has occurred within the state and, in
order to be consistent with in-state consumption, it
may reflect electricity imports and exports. It also ac-
counts for transmission and distribution losses.
Consumption and/or generation-related baseline data
can be obtained from many sources, including:
State energy offices and departments of transporta-
tion (Figure 2.2.5 provides an example of energy
consumption by fuel type data collected by the
state of Wisconsin Office of Energy Independence),
" Consumer energy use profiles by sector,
Utility Integrated Resource Planning (IRP) filings,
Public utility commissions,
Independent system operators (ISOs),
North American Electric Reliability Corporation
(NERC),
EPA's Emissions & Generation Resource Integrated
Database (eGRID),
DOE's Energy Information Administration (EIA),
and
DOE's National Renewable Energy Laboratory
(NREL).
As shown in Table 2.2.1, these sources provide a variety
of different types of data, including historical and pro-
jected supply and demand for electricity, natural gas,
and other fuels (discussed in the next section).
5 Local energy baselines can focus on end-use sectors (i.e., residential, com-
mercial, industrial, and transportation) and allocate the fuel used to generate
electricity across the sectors that consumed the electricity.
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 24
-------
STEP 1.3: Review and Select Method to
Forecast the Business-as-Usual Case
States can use basic or sophisticated modeling ap-
proaches to forecast their business-as-usual energy
cases and predict energy supply and demand. Both
approaches are based on expectations of future popula-
tion changes, energy data, and economics.
Basic methods may require a state to (1) adopt assump-
tions made by utilities, independent system operators,
and regulatory agencies about the projected popula-
tion, energy situation, and the economy; or (2) compile
and develop its own assumptions. Basic approaches are
generally appropriate when conducting screening anal-
yses or developing highly aggregated forecasts when
the amount of time or funding to support a forecast is
limited or when the time period of the forecast is short.
More sophisticated methods can be used for short-
term or long-term analyses. They provide greater detail
TABLE 2.2.1 SAMPLE ENERGY DATA SOURCES FOR DEVELOPING BASELINES AND BAU FORECASTS
Description
State Sources
Consumer energy
profiles (residential,
commercial, industrial)
Most utilities conduct audits, surveys, or EE evaluation studies as part
of energy efficiency programs' regular reporting. Data are customer-
specific load profiles that can be used to build up total demand.
State Energy, Utility
Commissions, Transportation,
or other Offices
Most states collect historical and forecast data for both supply and
demand information. Other agencies may have compiled similar
energy information that could be used for this effort.
Utility-Related Sources
Utilities
Most utilities collect historical and forecast data. Make sure
documentation is collected as well, so that limitations can be
understoodwhat's in and what's not, for example.
Consumer energy profiles
(residential, commercial,
industrial)
Most utilities conduct audits or EE evaluation studies as part of
energy efficiency programs' regular reporting. Data are customer-
specific load profiles that can be used to build up total demand.
Public Utility Commissions
Most PUCs collect historical and forecast data. Usually are supplied
from utilities and studies. Use to collect supply and demand data.
Independent System
Operators/ RTOs
Supply and total demand information to be used for planning
purposes. Available from the Midwest Independent System
Operator (ISO), ISO-New England, Pennsylvania-New Jersey
Maryland Interconnection, Southwest Power Pool, California ISO,
Electric Reliability Council of Texas, Florida Reliability Coordinating
Council, and New York Independent System Operator.
North American Electric
Reliability Corporation
(NERC) Electricity Supply and
Demand Database
Capacity and demand, up to 10-year projections of electricity
demand, electric generating capacity, and transmission line
mileage. Generation data include unit-level statistics on existing
generators, planned generator additions and retirements, and
proposed equipment modifications. Free to government agencies.
h ttp://www.nerc.com/page.php ?cid=4\38
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 25
-------
TABLE 2.2.1 SAMPLE ENERGY DATA SOURCES FOR DEVELOPING BASELINES AND BAU FORECAST fcontj
rtri; "r' ;
HHHHH
Sources 1 ฃ 1 ฃ 1 ฃ Description
Federal Agency Sources
EIA Electric Power Annual
EIA State Energy Profile, State
Energy Data (SEDS)
EIA Electric Sales, Revenue,
and Price tables or EIA Annual
Electric Utility data-EIA-860,
906, 861 data file
EIA Manufacturing Energy
Consumption Survey (MECS);
Commercial (CBECS);
Residential (RECS)
EIA Annual Energy Outlook
EPA Emissions & Generation
Resource Integrated
Database (eGRID)
NREL
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
National, some regional and state level capacity and demand,
margin, energy retail sales (MWh), revenue, emissions, short term
plans, etc.
http://www.eia.doe.gov/cneaf/electricity/epa/epa_sprdshts.html
Annual production, consumption, prices, and expenditures by
energy source.
h ttp://ton to. eia. doe.gov/sta te/
http://www.eia.doe.gov/cneaf/electricity/epm/tablel_6_a.html
http://www.eia.doe.gov/emeu/states/_seds.html
Annual data, peak, generation, demand/consumption, revenues,
utility type, and state.
http://www.eia.doe.gov/cneaf/electricity/epa/epa_sum.html
h ttp://www.eia.doe.gov/cneaf/electricity/page/eia861.html
http://www.eia.doe.gov/cneat Yelectricity/page/eia906_920.html
A national sample survey on the stock of U.S. buildings, their
energy-related building characteristics, consumption (by
appliance) and expenditures.
h ttp://www.eia.doe.gov/emeu/mecs/con tents.h tml
http://www.eia.doe.gov/emeu/cbecs/
http://www.eia.doe.gov/emeu/recs/contents.html
National forecast of supply and demand.
http://www.eia.doe.gov/oiaf/aeo/
http://www.epa.gov/egrid for supply planning.
Data on various renewable energy technologies and some costs.
h ftp://www.nrel.gov/rredc/
than the basic methods, and can capture the complex
interactions within the electricity and/or energy
system. Some states might want to consider a more
sophisticated modeling approach for their demand and
supply forecasts in cases where:
They want to better understand the effects of de-
mand growth on their required portfolio of supply
resources in the future, or
They want to analyze the effects on energy demand
and supply of significant changes that have oc-
curred or are expected to occur in economic pat-
terns (e.g., a dramatic decrease in housing starts)
or energy costs.
Sophisticated approaches are often data-, time-, and
labor-intensive; lack transparency; may involve model
licensing and data fees; and require a significant com-
mitment of staff resources to develop expertise in the
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 26
-------
TABLE 2.2.2 COMPARISON OF BASIC METHODS FOR FORECASTING ENERGY DEMAND AND SUPPLY
Methods
Compilation
of individual
forecasts by
others
Advantages
Easy to gather
Disadvantages
Driven by different assumptions that may no longer apply;
proprietary concerns; possible short horizons; may or may
not provide information on construction requirements,
fuel use, emissions, and costs; gaps in coverage.
High level, preliminary and
quick analysis.
Adoption of a
complete forecast
used by others
Easiest method May not have the long-term outlook.
Assumptions may not comport with desired state/
regional outlook.
May require translation to alternative geographic scope.
May be proprietary.
High level, preliminary and
quick analysis.
Nominal Group
Techniques (NGT)
Consensus
building
Time-consuming and relatively expensive.
Adequate budget and
stakeholder interest.
Linear and/
or Nonlinear
Extrapolation of
Baseline
Quick
May not capture impact of significant changes (e.g., plant
retirements).
High level with simple
escalation factors from history
or from other sources.
More robust data
analysis
Possible errors in formulas, inaccurate representation of
demand and supply.
Knowledge of generation
dispatch by type of plant.
model. Unless the tool is used for broader or multiple
analyses (e.g., statewide energy planning), it maybe
impractical for the state to build the capacity to run
these models in-house. However, most models are sup-
ported by one or more consultants who have readily
available supporting data and who may be retained for
these types of specialized studies.
This section provides information about basic and
sophisticated approaches, methods under each ap-
proach, data needs, and the respective advantages and
disadvantages of each of the methods.
Basic Forecast Methods: Demand and Supply
States can use a range of basic methods to project
their BAU energy without using rigorous, complicated
analyses and software models. These methods generally
produce aggregate information about a state's energy
future, perhaps with a larger margin of error than more
sophisticated approaches.
Basic approaches for forecasting energy demand and
supply include: (1) compilation of partial forecasts
(e.g., utility service territory) by others into one state
forecast; (2) adoption of a pre-existing forecast that
someone else may have developed for the state; (3)
group consensus-building processes to develop as-
sumptions used within a forecast; and (4) extrapolation
of historical rates of demand growth and electricity pro-
duction (or rates of growth from other forecasts) that
are applied to the baseline. Table 2.2.2 summarizes the
advantages and disadvantages of each approach and de-
scribes the most appropriate uses of these approaches.
Each approach is explained in greater detail below.
Compilation of individual forecasts by others: Ener-
gy plans from utilities, ISOs, and regulatory agen-
cies often include a demand forecast that reflects
energy savings from energy efficiency programs.
Similarly, a corresponding supply plan is likely to
include data on existing and projected renewable
energy sources, including combined heat and
power plants, if significant. States can aggregate in-
dividual load forecasts, generation expansion plans,
and energy efficiency programs and renewable en-
ergy evaluations from state agencies, utilities, ISOs,
local educational institutions, and special interest
groups, such as interveners in rate cases. Compil-
ing forecasts created by different entities can be
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 27
-------
challenging, because they can vary significantly
from each other in terms of underlying assump-
tions, proprietary concerns, data transparency (e.g.,
unit generation, costs), and time frame.
Adoption of a forecast used by others: In some
states, an energy office, utility commission, revenue
department, or academic organization may have
prepared a suitable energy forecast. Also, utilities
and ISOs may have forecast plans. A regulatory
filing requirement (e.g., Integrated Resource Plan)
typically provides a comprehensive long-term plan
that includes impacts from energy efficiency; reli-
able demand response, if any; and existing renew-
able energy plans.6 However, there may be propri-
etary constraints to obtaining this information and
these forecasts may reflect economic conditions
that differ from the state's view.
Nominal Group Techniques (NGTs) are structured
group processes (similar to "voting") to form
consensus opinions, including expectations for the
future. They can be used to develop forecasts or to
develop inputs to the preceding methods or more
complex models. The type most commonly used
in forecasting is the Delphi method. A more recent
approach, called Deliberative Polling, might be
useful for this purpose, but it is expensive and time-
consuming. Working with multiple stakeholders
does provide value overall; however, this approach
loses detail when valuing the impacts of changes.7
Linear-/Non Linear Extrapolation involves spread-
sheet analysis where historical demand growth
rates and electricity production trends (or trends
from an alternative forecast) are used to extrapo-
late base year data into the future. The accuracy
of this approach depends on the accuracy of the
"borrowed" growth rates, and the knowledge and
experience of the analyst when applying histori-
cal trends. An advantage to this approach is that
it is easy to develop in a spreadsheet and use for
preliminary forecasting. A disadvantage is that
the exclusion of important variables beyond
demand growth factors and electricitysuch as
weather; season; plant retirements or construction,
6 For information about how utilities integrate energy efficiency into resource
planning, see The Guide to Resource Planning with Energy Efficiency: A
Resource of the National Action Plan for Energy Efficiency, November 2007.
www.epa.gov/cleanenergy/documents/resource planning.pdf
7 In Vermont, a similar approach was used through a public workshop pro-
cess in which electric industry stakeholders provided their input on the state's
energyplan.
FIGURE 2.2.6 NEW JERSEY ENERGY PLAN-
BASIC DEMAND FORECAST
Projected Electricity Growth Rate in New Jersey for all Sectors
2005-2020
2D05 2C06 2007 2008 2039 2010 2011 2D12 2C13 2014 2015 2016 2017 2018 2019 2020
Year
jAII Sectors (1.52% Annual Growtn)
This BAU electricity forecast was developed using a relatively
simple approach in which past load growth rates were reviewed
and assumptions were made regarding the ways in which
industry trends and existing policies affect future growth
patterns. While recent growth rates (1998 to 2004) had been in
the range of 2% annually, the average annual growth rate since
1990 was only 1.52%. The New Jersey Board of Public utilities
chose to carry forward the long term 1.52% growth rate based
on the assumption that demand growth would level out once
electricity prices increase after the deregulation rate caps expire.
Source: New Jersey, 2008.
operation, or capital costs; emissions; or macro-
economic growthmay result in an inaccurate
forecast. Figure 2.2.6 illustrates a simple example of
a linear extrapolation analysis.
Sophisticated Forecast Methods
States may develop supply and demand forecasts using
one of the basic approaches described above, based
on the perception that the demand rate will probably
follow historical trends. Alternatively, they might want
to consider a more sophisticated modeling approach
when they require a more comprehensive understand-
ing of their energy profile or when they have experi-
enced or expect to experience significant changes in
their energy or economic patterns.
Sophisticated methods involve data- and resource-
intensive computer-based models that generate
detailed forecasts that may reflect historical trends,
economic and/or engineering relationships, future
expectations about prices, technologies and technology
development, operating constraints, and regulatory
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 28
-------
expectations (e.g., environmental regulations). While
basic forecast methods are applied similarly to demand
and supply forecasts, sophisticated approaches gener-
ate separate demand and supply forecasts that can
be integrated once developed. As such, sophisticated
methods for developing demand and supply forecasts
are described separately below.
Demand Forecast
Once the historical baseline is developed, states can
develop an energy demand forecast using time series,
end use, or econometric models. These types of models
can be used for short- and long-term load forecasting,
comprehensive load analysis, modeling, and "day-after"
settlement. Each model, and its advantages and disad-
vantages, is described below.
Time-Series Models are based on the assumption
that the data (and the variable being forecast) have
a structure or pattern, such as a trend and/or sea-
sonal variation. Future events are forecast based on
known past events and patterns. Inputs require an
analysis of historical patterns in demand for elec-
tricity. This analysis can be a simple look at the ag-
gregate demand and a forecast based on the pattern
of this demand, or a breakdown of the demand
into customer type (e.g., residential, commercial,
industrial) and application of each cyclical pattern
over time to develop the total demand forecast.
Advantages of time-series models are:
These models are easy and fast to use; and
> Historical data are widely available by year,
fuel, end use, or sector (residential, commer-
cial, and industrial).
Disadvantages of time-series models are:
> Data may relate to a historical baseline that
has undergone major structural changes,
such as a switch from heavy manufacturing to
high-technology industries, that are unlikely to
occur again, thus complicating or invalidating
the forecast;
> It is hard to reflect future structural changes
even if they are anticipated; and
Time-series models cannot reflect supply-
demand-price feedbacks dynamically.
End-Use Models develop the load profiles of each
customer type by analyzing the historical con-
sumption of appliances and equipment (including
any existing DSM programs) and may use specific
surveys from customers about future growth and
contraction. This approach can also include an
economic forecast that provides gross state product
(GSP) and consumer electricity prices.
An advantage is that this approach uses a load
profile for each customer class being served,
providing a reasonable estimate of demand.
A disadvantage is that it can require consider-
able time and cost to collect the data. Users
can elect to use project-specific models to help
assess building demand estimates.
Econometric Models provide a more complex and
robust analysis that uses inputs such as inflation,
demographics, gross state product, consumer ener-
gy prices, gross/disposable income, housing starts,
business starts/failures, birth/death rates, surveys
of business expansion plans, historical energy
consumption, and other variables for structural
changes and economic data. The model output in-
cludes data correlations, or relationships, between
demand and energy consumption. For example, the
output may show that as income increases, energy
demand increases. These relationships can be ap-
plied in detailed demand and energy consumption
forecasting. Econometric methods are sometimes
used in combination with end-use methods.
An advantage of this method is that it creates a
robust demand forecast if driven with a robust
economic forecast.
A disadvantage is the time and cost required to
prepare the inputs and review the results.
Some examples of these models in use include
ENERGY 2020 and EPRI's suite of tools. ENERGY
2020 is an end-use-econometric energy market
model used for forecasting demand and supply
across all fuels and sectors. It has been used in
Illinois, Massachusetts, and Hawaii for long-term
forecasting. EPRI's suite of bottom-up, end use
forecasting models, such as the Residential End
Use Energy Planning System (REEPS) and the
Commercial End Use Planning System, are used
primarily by utilities. Some states have developed
their own models. For example, California has
developed end use (residential) and econometric
(commercial) models for forecasting.
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 29
-------
Supply Forecast
Utilities, ISOs, and other sophisticated energy market
participants use supply forecast models for hourly,
daily, monthly, and long-term forecasting. Sophisti-
cated supply forecasting models require large volumes
of data on electricity production plants, transmission
capabilities, and a demand forecastand the better the
quality of that data, the better the results. Although the
costs to acquire the software and data may be prohibi-
tive for some users, these models generally provide
more robust estimates on energy and capacity output
than basic modeling approaches. Models covering both
electricity dispatch modeling and capacity expansion
(or planning) modeling are summarized in Table 2.2.3.
Electricity Dispatch models (also commonly re-
ferred to as "production cost" models) simulate the
dynamic operation of the electric system, generally
on a least-cost system dispatch. In general, these
models optimize the dispatch of the system based
on the variable costs of each resource and any
operational constraints that have been entered into
the model. These models are helpful in assessing
which existing plants8 are displaced. These models
8 These dispatch, or production costing models focus on existing plants or a
specified portfolio of plants (which may contain some new or proposed plants);
however, these models only produce estimates of avoided variable costs and
changes in the output of different resources. Changes in the use of a resource
(e.g., a marginal coal-fired power plant) are key inputs into any modeling of
changes in emissions due to EE or RE activities. These dispatch models do not
internally examine changes in the capital costs (e.g., avoided capital costs) that
might result from investments in EE or RE. However, this can he done through
spreadsheet models that have been developed to augment electricity dispatch
models or using models that combine capacity expansion and dispatch (e.g.,
NEMS, IPM).
TABLE 2.2.3 EXAMPLES OF SOPHISTICATED SUPPLY FORECASTING MODELS
Sampling of models
Electricity Dispatch
Advantages
Disadvantages
When to Use this Method
PROSYM
GE MAPS
PROMOD IVฎ
MIDAS'
Can provide very detailed
estimates of specific plant
and plant-type effects within
the electric sector.
i Provides highly detailed,
geographically specific,
hourly data.
Often lacks transparency.
Labor- and time- intensive.
Often high labor and
software licensing costs.
Requires establishment of
specific operational profile of
the clean energy resource.
Often used for evaluating:
Specific projects in small
geographic areas.
Short-term planning (0-5
years), and
Regulatory proceedings.
Capacity Expansion or Planning
NEMS
IPMฎ
ENERGY 2020
LEAP
Strategistฎ
Plexosฎ
EGEAS
AURORAxmp
MARKAL-MACROb
Ventyx System Optimizer
Model selects optimal changes
to the resource mix based on
energy system infrastructure.
May capture the complex
interactions and feedbacks
that occur among demand,
environmental, fuel, electric
markets.
Provides estimates of emission
reductions from changes to
the electricity production and/
or capacity mix.
May provide unit-specific
detail (IPM).
Requires assumptions that
have large impact on outputs.
May require significant
technical experience.
Often lacks transparency.
Labor- and time- intensive.
Often high labor and
software licensing costs.
Long-term studies (5-
25 years) over large
geographical areas, such as:
State Implementation Plans,
Late-stage resource planning.
Statewide energy plans, and
Greenhouse gas mitigation
plans.
Ventyx markets the MIDAS solution as a strategic planning tool since it incorporates Monte Carlo capabilities. This tool is included in the
list of electricity dispatch models, as it generally uses a pre-selected set of resource plans and the MIDAS model focuses on electricity
price forecasting and financial analyses (e.g., balance sheet analyses) of each resource plan.
MARKAL-MACRO model is represented as multipurpose energy planning model, http://www.etsap.org/Tools/MARKAL.htm.
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 30
-------
are also used in short-term planning and regula-
tory support.
Advantages of electricity dispatch models are:
These models are often used in generation
project financial analyses, since they provide
forecasts of wholesale electric prices for each
hour (i.e., system marginal costs) and the
hourly operations of each unit. By comparing
the variable costs of each unit with the price
forecasts, an estimate of plant profitability can
be developed.
> They can be run to develop a BAU and mul-
tiple sensitivity cases to assess the impact on
various planning parameters (e.g., transmis-
sion, plant dispatch, and avoided variable
costs), and may capture complex interactions
and tradeoffs.
Electricity dispatch models are usually more
detailed in their specification of operational
and variable costs compared with capacity
expansion models.
These models are ideally suited for estimating
wholesale electric prices (i.e., the marginal
system cost) and the hours of operation and
production of each unit in the system for up
to a five-year time frame. This information has
been the basis for plant financing decisions
and the development of unit operating and bid
strategies in markets. In these roles, the elec-
tricity dispatch model is viewed as a superior
tool. These same data also are necessary in
estimating the emissions of specific units and
the regional electric system being modeled.
Disadvantages of electricity dispatch models are:
> These models cannot estimate avoided capacity
costs from EE or RE investments. Unlike the
capacity expansion models described below,
these costs must be calculated outside the
electricity dispatch model using a spreadsheet
model or other calculations.
>Some of these models require substantial detail
on each unit in a regional electric system and
are typically full chronologic models (i.e., some
data elements are needed for all 8,760 hours in
a year).
The complexity of these models often results
in agencies and stakeholders working with
utilities to coordinate the application of the
models in policy analyses and in regulatory
proceedings.
Electricity dispatch models can also be ef-
fectively used to develop estimates of genera-
tion impacts of long-term resource plans, but
they require considerable side calculations in
terms of the explicit specification of projected
new units that constitute a limited number of
"build scenarios" and the computation of the
capital costs of the system to augment the vari-
able costs produced internally by the electricity
dispatch models.
Capacity Expansion or Planning models are
designed to make decisions on how the electric
system adds new capacity to meet future demand
over a 20- to 25-year planning period. This differs
from the primary role of electricity dispatch mod-
els, which is to develop electricity price forecasts,
the hours of operation, the electricity output for
specific units, and the revenues and profits for
generation units in a regional system. In contrast,
capacity expansion models evaluate the economics
of potential new generating unit additions to the
system (some models allowing a great deal of spec-
ificity with respect to new unit options). Capacity
expansion models use information on demand
growth, regional electric system operations, and
the characteristics of candidate new units, typically
within an optimization framework, that selects
a future build-out of the system (multiple new
units over a 20- to 25-year time frame) that has
the lowest overall net present value (NPV), taking
into account both capacity and variable costs of
each unit. This simulated build-out can include the
retirement of existing units, selection of base load
capacity, and decisions to build peaking capacity
that minimizes the NPV over the 20- to 25-year
planning scenario.
Many capacity expansion models have some rep-
resentation of system dispatch. Dispatch modeling
in these combined capacity expansion and dispatch
models may not be based on an 8,760 hourly
structure, but instead dispatch to more aggregated
load segment curves representing seasonal energy
demand by load segments (e.g., peak, intermediate
segments, and base load). These types of models
include IPM* and NEMS.
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 31
-------
Advantages of capacity expansion models are:
They are designed to incorporate a number of
factors that are influenced by changing poli-
cies, regulatory regimes, or market dynamics
(e.g., stricter emission policy, introduction of a
renewable portfolio standard).
> While both electricity dispatch models and
capacity expansion models are used in IRP
proceedings, the capacity planning model is
designed specifically to develop long-term
resource plans.
> Capacity expansion models are able to estimate
avoided capacity costs and usually also pro-
duce estimates of avoided variable costs.
Disadvantages of capacity expansion models are:
The complexity of these models often results
in agencies and stakeholders working with
utilities to coordinate the application of these
models in policy analyses and in regulatory
proceedings.
STEP 1.4: Determine Assumptions and
Review Data
After choosing the forecasting approach or model type,
the next step is to determine or review assumptions
about population, energy, and economic variables,
such as energy prices, productivity, gross state product,
and the labor force upon which projections of energy
demand and supply depend.
It is also important to review possible data sources and
collect the data required for the analysis. The following
types of data are used in estimating energy consump-
tion and supply baselines and forecasts:
States can use population data to estimate the
amount and types of demand expected in the
future and to examine trends. The U.S. Census
Population Estimates Program provides historical
and projected population data (http://www.census.
gov/popest/estimates.php).
* A forecast depends upon assumptions about the
economy that the analyst projects into the future.
States can examine economic variables as they
relate to energy in order to better understand the
historical relationships between energy and the
economy, and to anticipate how these relationships
may exist in the future. The Bureau of Economic
Analysis (http://www.bea.gov/), Bureau of Labor
and Statistics (http://www.bls.govf), and the U.S.
Census Economic Census (http://www.census.gov/
econ/census02f) all provide macroeconomic data
that states can use.
The forecast may require assumptions about the
energy and fuel prices the state should expect in
the future. EIA provides regional energy and fuel
price forecasts out to 2030 (http://www.eia.doe.
gov/oiaf/forecasting.html). Price projections may
also be available from PUCs and ISOs, although
proprietary constraints may limit the amount
available. In addition, a number of private data
providers may be able to offer data that are more
recent than those from publicly available sources.
Almost all providers of electricity dispatch and capacity
expansion models also offer a data set that can be used
to apply these models to a regional electric system.
Data from any source must be examined to ensure that
they are consistent with the assumptions of the entities
that will use the model results, and to check for outli-
ers, errors, and inconsistencies in the data. No data set
from any source is guaranteed to be fully appropriate to
a user's needs, and any data set may contain errors.
At this point in the process, it may also be necessary to
clean the data and/or fill in any missing data gaps. If data
points are missing for particular years, it may be neces-
sary to interpolate the existing data or use judgment to
fill in gaps. This will minimize the likelihood of generat-
ing results based on calculations that are skewed due to
missing or out-of-range data, producing a forecast that
would then not make sense. Some of the private data
providers also provide data cleaning services. Practical
application of any of these data bases, however, requires
due diligence in looking for data outliers, missing values,
and screening for errors in data. It is a rare occurrence
for a user to obtain a fully clean data set, consistent with
their individual assumptions, from any one source.
STEP 1.5: Apply Model or Approach
States can apply the selected model or approach to the
historical baseline energy data based on the assump-
tions about future population, economic, and energy ex-
pectations. It is important to revisit the assumptions and
data that will be required for the specific model require-
ments to assure that they are still valid. As mentioned
in earlier sections, many state agencies and stakeholders
work with utilities or consultants to actually perform
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 32
-------
the model runs. Still, it is important to have transpar-
ency around the model inputs and the policy/regulatory
assumptions incorporated into the model, as well as a
solid understanding of the basic operations of the model
(i.e., the algorithms used to produce the model outputs).
STEP 1.6: Evaluate Forecast Output
Once generated, it is important to evaluate the forecast
to ensure that it is reasonable and meets the original
objectives. If the state determines that some or the
entire forecast does not seem realistic, it may need to
revisit assumptions and then re-apply the approach or
model to achieve an acceptable demand forecast.
Issues and Considerations
When developing an energy baseline and BAU forecast,
it is important to consider the following issues.
Typically the data available for a baseline and
BAU forecast lag several years. For this reason,
the current and most recent years may be part of
the forecast and not the history. It is important,
therefore, to ensure that the data derived for recent
years reflect the current energy supply and demand
as much as possible.
As with all analyses, transparency increases
credibility. All sources and assumptions require
documentation.
When documenting an energy forecast, it is
important to clearly state what activities will take
place without any new clean energy initiatives (i.e.
what is "in the baseline"). For example, many state
forecasts assume that some level of energy-efficient
actions or regulatory changes (e.g., GHG reduction
requirements) will be implemented over time. It is
important to avoid double-counting when examin-
ing future program potential or impacts.
2.2.2 STEP 2: QUANTIFY IMPLICATIONS
OF TARGETS AND GOALS
If a state has or is considering a broad clean energy
goal, it is helpful to estimate the potential implications
of the goal before evaluating specific clean energy
programs and implementation options. For example,
the state may need to quantifyin terms of kWhsthe
requirements of an energy efficiency goal or target.
Suppose the policy or goal is to have zero growth in en-
ergy demand over the next 10-20 years; it would then
be necessary to estimate how much energy efficiency
would be required to meet that goal. Alternatively, the
state may need to quantifyagain, in kWh termsthe
implications of a renewable portfolio standard. These
estimates will indicate how much energy must be saved
each year, or how much clean energy must be provided.
While the energy implications of any goals should be
checked against existing energy efficiency or renewable
energy potential studies to make sure they are plau-
sible, this type of estimate is not focused on estimating
what is cost-effective, what the market might adopt,
or when the specific technologies might be adopted; it
only estimates what the goal or target implies.
Methods for these estimates can include both basic and
sophisticated approaches, but these high-level estimates
will most likely require only the most basic approaches
as the focus is simply on quantifying the meaning of
the goal (e.g., a 2 percent reduction in demand per year
implies a savings of x kWh). Basic approaches typically
start with a baseline forecast as developed under Step 1.
This will be the primary determinant of energy savings
or clean energy supply required. The exact methodol-
ogy chosen, however, will depend on how the goal or
target is specified and a host of other factors, such as
whether the energy savings from efficiency are mea-
sured from the baseline forecast or from prior years'
sales. Also, the extent to which existing programs do or
do not count toward the target may affect the calcula-
tions. It is important to read (to the extent they are
available) the details of the goal, policy, or legislation,
then think through the implications of these details for
the methodology and calculations.
Suppose a state is determining the anticipated energy
savings or generation needed to achieve a clean energy
initiative in a target year (e.g., the target is to build 100
MW of wind power capacity by 2020). If appropriate
financial incentives are in place to encourage construc-
tion of the wind facility, the energy available in the year
after 100 MW of wind facilities are placed in service
can be estimated at a very basic level as:
100 MW * 0.28 capacity factor9 * 8,760 hours/year =
245,280 MWh/year.
The important element here would be to ensure that
the 28 percent capacity factor is applicable to the
9 Capacity factor is defined as the ratio of the electrical energy produced by
a generating unit for the period of time considered to the electrical energy that
could have been produced at continuous full power operation during the same
period. Typical capacity factors for wind range from 20 percent to 35 percent.
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 33
-------
EXAMPLES OF STATE ENERGY TARGETS OR GOALS
Have a rate of zero load growth by 2020.
Reduce electricity demand by 2% per year by 2015, and 2%
every year thereafter, with reductions to be based on prior
three years' actual sales.
Meet 20% of generation requirements (or sales) through
renewable energy sources by some date in the future
(sometimes with interim targets). In some instances, the
eligible resource types (including existing), the required
mix of renewables types, and geographic source of the re-
newables may be specified.
wind resource being considered. The output of a wind
turbine depends on the turbine's size and the wind's
speed through the rotor, but also on the site's average
wind speed and how often it blows. Data to assess ap-
propriate capacity factors can be identified based on
geographic data on wind class (speed).
Alternatively, suppose a state is considering an Energy
Efficiency Portfolio Standard (EEPS) that calls for a 20
percent reduction in energy demand growth by 2020.
The state might estimate the annual implications of the
policy as outlined below (with calculations illustrated
in Table 2.2.4).
First, a pathway, with annual targets, would be
required to assure the 20 percent total reduction is
reached. Table 2.2.4 shows one possible pathway.
Next, this percent savings is applied to the BAU
forecast (which was expected to increase by 3 per-
cent per year prior to the EE initiative) in order to
calculate EE savings required. The fourth column
shows the EE savings required.
Finally, the new target level of demand is shown.
In this example, the results indicate a new lower
demand annual average growth rate (AAGR) of
1.1 percent.
While the actual path that is followed or the estimates
of achieved savings (e.g., for M&V purposes) may dif-
fer from those shown in this simple exercise, this type
of calculation gives an indication of the implications
for program requirements and the resulting impact on
growth.
TABLE 2.2.4 EXAMPLE OF ESTIMATION OF REQUIRED EE SAVINGS BASED ON LONG TERM SAVINGS
GOAL OR PERFORMANCE STANDARD (KWH)
BAU Demand (3% AAGR) BAU required
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
New Target Demand (New
EE Savings Required AAGR = 1.1%)
1,000.0 0.5% 5.0 995.0
1,030.0 1.0% 10
1,060.9 1.5% 15
.3 1,019.7
.9 1,045.0
1,092.7 3.5% 38.2 1,054.5
1,125.5 5.5% 61.9 1,063.6
1,159.3 7.5% 86.9 1,072.3
1,194.1 9.5% 11
1,229.9 11.5% 14
3.4 1,080.6
1.4 1,088.4
1,266.8 13.5% 171.0 1,095.8
1,304.8 15.5% 202.2 1,102.5
2018 1,343.9 17.5% 235.2 1,108.7
2019
1,384.2 19.5% 26
9.9 1,114.3
2020 1,425.8 20.0% 285.2 1,140.6
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 34
-------
If the state has an emissions-related goal, this type of
quick, top-down analysis can then be linked to emis-
sions data to determine what portion of the state's
emissions targets could be met with a specific percent-
age EEPS. Similar linkages could be made to economic
or other impacts as well.
Considerations
There are a number of factors to consider when
estimating the implications of targets and goals for
electricity demand and resources:
The baseline level of electricity demand and supply
(described earlier in this chapter);
Expected growth over time under BAU (described
earlier), including any ongoing energy efficiency or
renewable energy efforts that may or may not con-
tribute to the new goal, but will influence baseline
conditions;
The likely persistence of energy efficiency sav-
ings over time (or changes in the supply of clean
energy);
Other considerations that may affect the level of
savings or supply required, such as rebound effects
in energy efficiency programs; and
The remaining electricity demands (or supply)
after the impacts occur.
Quantifying the implications of broad goals and
targets typically requires straightforward mathemati-
cal calculations, as shown above, and do not usually
involve sophisticated approaches. However, advanced
modeling and economic analysis may be required if,
for example, a goal or target is tied in some way to an
economic indicator or requirement (e.g., if a goal or
target has some circuit-breaker or threshold provision,
for example, requiring that only energy efficiency
costing less than a certain amount be required), or has
some dynamic aspects to it (e.g., changing targets in
response to achievements).
2.2.3 STEP 3: ESTIMATE POTENTIAL DIRECT
ENERGY IMPACTS
A critical step in the process of assessing the multiple
benefits of clean energy is the estimation of the poten-
tial direct energy impacts of clean energy programs or
policies under consideration. Direct energy impacts in-
clude energy savings from energy efficiency initiatives
PROGRAMS FOR WHICH ENERGY IMPACTS
MIGHT BE ESTIMATED
Energy Efficiency Portfolio Standards
Renewable Portfolio Standards
Appliance Standards
Building Codes
Public benefits funds (to fund state or utility-run efficiency
or renewables)
Clean Energy Tax or other Financial Incentives
Rebate programs
Lead by Example Programs
and electricity production from renewables and other
clean energy supply options. These estimates are the
foundation for estimating the multiple benefits of clean
energy as described in the subsequent chapters of this
Resource. For example, changes in energy consumption
due to energy efficiency or energy output from clean
resources are matched to characteristics of generation,
as described in Chapters 3 and 4, to assess changes in
costs, emissions, and other factors.
Potential direct energy impact estimates can be de-
veloped in the context of a target, but a target is not
required to estimate these impacts. Here the state
would be estimating the expected result of a policy
or program that is under consideration and has been
sufficiently defined to allow meaningful analysis. In the
case of prospective programs and policies, the state is
trying to assess whether the program or policy goals
are achievable and at what costs, and what specific ac-
tions are required by market participants. For example,
the state may be considering an RPS of 20 percent by
the year 2020, and wants to understand what specific
resources would have to be built to comply; or the state
may have a goal of 10 percent reduction in residential
energy demand in five years and wants to understand
what programs it can implement to achieve that goal.
Examples of these types of impact estimates include:
Estimating the impact of appliance standards in a
way that considers the existing stock, current ef-
ficiency levels, and consumer decision making;
Estimating the expected response to a utility en-
ergy efficiency program, with or without specific
information on program focus (what sectors and
end uses) and design issues (e.g., rebate levels); and
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 35
-------
Estimating the impact of a renewables incentive
program.
Please see the text box Programs For Which Energy
Impacts Might Be Estimated for program examples.
Similar to the process for developing an energy fore-
cast, estimating the potential direct energy impacts
involves a series of steps, including:
1. Define Objectives and Parameters,
2. Choose Method to Estimate Potential Direct En-
ergy Impacts,
3. Determine Assumptions and Review Available
Data,
4. Apply Model or Approach, and
5. Evaluate Output.
Each of the steps is described in greater detail below.
STEP 3.1: Define Objectives and Parameters
It is important to define the objectives and parameters
of the direct energy impacts a state plans to estimate.
If the objective is to quantify the required energy sav-
ings from a state's clean energy initiatives or goals to
the state legislature, for example, the parameters of
the analysis may already be dictated. For example, the
legislature has likely specified a due date, a time period
to be analyzed, and a reasonable level of rigor, and may
even have required the state to spend a certain amount
of money on the analysis. Other analyses, such as those
conducted to screen a range of clean energy options
based on their multiple benefits, may be less defined.
It is necessary to consider each of the following param-
eters before choosing an analysis method, model, or
dataset(s) to use.
Time period for the direct energy impacts: Is it a
short-term or longer-term projection?
Timeliness of the estimates: Is this due in a year or
next week?
Level of rigor necessary to analyze policy impacts: Is
this for a screening study or a regulatory analysis
that is likely to be heavily scrutinized?
Availability of financial, staff, and outside resources
to complete the analysis in the required time period:
Is there a budget available for the analysis? Does
the state have internal modeling capabilities?
Amount of data available, or that can readily be
acquired, to develop the savings estimate: Are there
existing clean energy potential studies or similar
projects elsewhere that can be adapted to a state
analysis?
These factors will help states choose between simple
and more rigorous approaches based upon specific
needs and circumstances.
STEP 3.2: Choose Method to Estimate
Potential Direct Energy Impacts: Energy
Savings and Renewable Energy Generation
Several tools and methods are available to help states
estimate the potential direct energy impacts of clean
energy options. States can conduct their own surveys
or studies to estimate the direct energy impacts of clean
energy policies and use sophisticated methods, such as
applying building simulation tools, vintaging models,
and production costing models. Because new surveys
and studies tend to be costly and time-consuming,
however, states often use those that have already been
done by utilities, trade groups, other states, or the
federal government, and adapt them to reflect the cir-
cumstances of the state. It is likely that states will need
to use a combination of both existing and new analy-
ses, since existing data sources and studies must be
supplemented with complete and up-to-date data for
specific populations and measures that can be difficult
to obtain without additional targeted research.
Estimates typically factor in several considerations,
including:
the characteristics of the customer base and the
existing equipment stock,
the economics of the clean energy options and
their alternatives, and
the behavior of the market.
For example, to understand the generation system im-
pact of renewable energy resources, it is important to
understand not only how much renewable energy is re-
quired to meet the policy and therefore is coming into
the grid, but what type of renewable resource will be
available and that resource's operating characteristics
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 36
-------
(capacity factor, energy generation profile).10 States also
want to understand the cost and other impacts of the
energy efficiency and renewables driven by the clean
energy policy or mandates.
These types of questions require methodologies and
approaches that consider technology characteristics,
economics, and market conditions. For example,
estimating energy impacts from energy efficiency re-
quires an understanding of the current penetration of
a technology, applicability to new (or existing) homes,
customer financial requirements and preferences,
penetration patterns, and load shape impacts. Analysis
of appliance standards or building codes requires
understanding the technologies, but also the system
impacts at the building level. In addition states must
understand the potential impacts across the entire
population of affected buildings. Again, more advanced
techniques may be required, such as building simula-
tion tools and market penetration models, but some
basic non-modeling methods may apply. The range of
approaches is described below.
Approaches
Assessing the potential impacts of energy efficiency
or renewable energy programs requires "bottom
up" economic and/or engineering-based estimation
techniquesbuilding up estimates of impacts based on
a representation of the fundamentals of the technology,
the economics, and market behavior. These bottom-up
approaches involve estimating potential energy savings
at a very detailed level and rolling these estimates up to
the clean energy or statewide initiative level.
Analyses typically involve basic to sophisticated cal-
culations or spreadsheet analysis, and the collection of
data and information about the experiences or analyses
of programs within and outside of the state. Depending
upon the level of sophistication used in the analysis,
the analysis may or may not consider explicitly local
economics, transmission requirements, or generation
system impacts. The most basic types of analyses (i.e.,
those that exclude those factors) may be useful only for
developing short-term impact estimates, depending on
the extent of the comparable historical experience.
Depending upon the level of detail desired and the
amount of new analyses needed, estimating the
10 For information to help a state decide ifbiomass is a viable renewable
energy option to consider and, ijso, the most promising options to pursue,
see EPAs State Bioenergy Primer http://www.epa.gov/statelocalclimate/
resources/bioenergy-primer.html
potential impacts can require an extensive amount of
data and, for the more detailed analyses, may be costly.
At a minimum, the analysis will require some level of
detail about the:
Individual measure savings or renewable energy
savings that can be rolled up into an aggregate esti-
mate or state-wide strategy, and
Saturation of energy efficiency or renewable energy
equipment in the market so that the state can deter-
mine how much opportunity for new investment is
feasible when compared against potential studies.
Individual Measure or Site-level Savings
for Generation Estimates
To estimate the potential savings of clean energy mea-
sures, states can conduct simple analysis of estimated
energy efficiency or renewable energy impacts based
on an extrapolation of existing energy efficiency or
renewable energy potential studies. These studies may
be sector-specific (residential, commercial, industrial),
or more aggregated at some geographic level (state or
region). They may reflect technical potential, economic
potential, or market potential, or all three. If only the
first two estimates are provided, the analysis should
consider what is achievable.11
States can also explore existing studies of similar
programs in other states and adapt the results to their
conditions. At the aggregate level this may mean scal-
ing results to the state's load forecast, perhaps account-
ing for sectoral share differences if data are available at
the sectoral level. For estimates of individual measure
impacts or site-level impacts associated with clean
energy measures, states can look to available retrospec-
tive studies that can be extrapolated into prospective
savings based on an understanding of the state's
sectoral and end-use mix. Table 2.2.5 lists resources on
retrospective savings estimates and existing potential
studies states can use to produce individual savings
estimates.
These estimates can be summed across the populations
in each sector, remembering to subtract the market
penetration levels for the clean energy measures that
are already installed (based on the saturation data, as
11 EPA has developed guidance on conducting an energy efficiency potential
study. See Guide for Conducting an Energy Efficiency Potential Study, A
Resource of the National Action Plan for Energy Efficiency, November 2007.
http://www.epa.gov/RDEE/documents/potential^uide.pdf
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 37
-------
TABLE 2.2.5 RESOURCES FOR CLEAN ENERGY RETROSPECTIVE DATA AND POTENTIAL STUDIES
Description
Market Assessment and
Program Evaluation
(MAPE) Clearinghouse
Database developed by Consortium for Energy Efficiency (CEE) that
contains energy-efficiency program evaluation reports, potential
studies, and related documents that are publicly available.
Web Site
http://www.ceel.org/eval/
clearinghouse.php3
Lawrence Berkeley
National Laboratory (LBL)
Technical resource that tests and invents energy-efficient
technologies and provides publicly available research reports and case
studies on EE and RE.
http://www.lbl.gov; http://eetd.
lbl.gov/ea/ems/cases/
Renewable Energy Policy
Project (REPA)
Research papers, primarily on RE. Example reports are "Wind Energy
For Electric Power" and "Powering the South: A Clean and Affordable
Energy Plan for Southern United States," which includes EE and RE.
h ftp://www.repp. org/repp/
American Council on
Energy Efficient Economy
(ACEEE)
Consumer resources on appliances, policy, potential study workshops,
technical papers.
h ttp://www.aceee. org/
Tellus Institute
High-level reports presenting scenarios on increased efficiency
and renewable energy standards, reporting on their impact on the
environment. Also provides additional links to the software models
used by the Institute, including LEAP (Long-range Energy Planning).
h ttp://www. tellus. org/
National Renewable
Energy Laboratory (NREL)
Provides data on RE and EE technology, market, benefits, costs, and
other energy information.
http://www.nrel.gov/analysis/
California Database
of Energy Efficiency
Resources (DEER)
Provides documented estimates of energy and peak demand savings
values, costs, and effective useful life. In this California Energy
Commission and California Public Utilities Commission sponsored
database, data are easy to research and could be used as input into
internally developed spreadsheets on appliances and other EE measures,
which can be adjusted for the circumstances of different states.
http://www.energy.ca.gov/
deer/
Regional Technical Forum
(RTF) deemed savings
database
Developed by the Northwest Planning Council staff, with input from
other members of the regional technical forum, which includes
utilities in the four-state region of Oregon, Washington, Idaho, and
Montana. Both residential and commercial EE measures are included.
h ttp://www.n wcouncil.org/
energy/rtf/supportingdata/
default.htm
Entergy Texas Deemed
Savings
Entergy, an investor-owned utility (IOU), provides deemed energy
savings for EE measures, much as the other lOUs in Texas do. It
accounts for the weather zone of the participants. These data could
be used as input into internally developed spreadsheet regarding
appliances and other EE measures for a bottom-up method. The data
may have to be adjusted for a different state.
h ttp://www.en tergy- texas. com/
content/Energy_Efficiency/
documen ts/HelperApplica tion_
HTR_Entergy_2006.xls
described in greater detail below). When implementing
this approach of adapting existing studies to evaluate
renewable energy options, states should correct for
the relative resource base available since states have
different levels of renewable energy resources (e.g.,
wind, solar) available. The results should be adjusted
to reflect any difference.
Saturation of Energy Efficiency or
Renewable Energy Equipment
It is important to understand how much equipment
is already in the market so that states can determine
a feasible level of investment that a new clean energy
program or policy could induce. The equipment satu-
ration data are typically determined using one or more
methods, including:
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 38
-------
End-use Customer Saturation Surveys. These sur-
veys provide a relatively cost-effective method of
estimating saturation levels for both standard and
efficient equipment. These on-site, telephone or
Internet surveys are conducted to gather informa-
tion regarding the end-use equipment currently
installed at a statistical sample of homes and
businesses.
Site Visits. Facility managers can provide high-
quality estimates of equipment saturations.
However, due to the tremendous amount of energy
consumption represented by large nonresidential
facilities, and the limited amount of program audit
data available, it is often necessary to conduct
primary data collection at a sample of sites that
represent the sub-sectors in the population.
Survey of Retailers. Retailers can provide important
insight into the market share and saturation of a
number of products, including programmable
thermostats, water heaters, clothes washers, clothes
dryers, and refrigerators.
Surveys of Builders and Code Officials, Builders, Ar-
chitectural and Engineering Firms, and Other Trade
Allies. These data can be also be used to characterize
the equipment saturations in the new construction
and retrofit markets if samples are carefully selected
and appropriate survey instruments developed.
Interviews with contractors, dealers, distributors,
and other trade allies provide a cost-effective
research approach, as business activity tends to be
concentrated among relatively few market actors.
Trade ally interviews can also be leveraged to assess
market share and estimates of market saturation for
multiple sectors during a single interview.
Once equipment saturation is understood, states can
compare it against potential studies to determine the
feasible level of investment opportunity available.
Tools for Direct Savings or Generation Estimates
A number of modeling and analytics tools are avail-
able to help states estimate the potential direct energy
impacts of clean energy measures. Table 2.2.6 provides
examples of some simple analysis tools available when
employing non-integrated modeling approaches to
estimating energy savings from EE and RE initia-
tives. The tools shown in the table are organized by
web-based, spreadsheet, and software tools. Some of
these tools are designed to develop site-level savings
estimates that can be aggregated up to the state.
For example, the site-level estimates from tools such as
eQuest* (for EE measures) or PVWatts (for estimating
solar system electricity production) are summed across
expected participant populations to get statewide
energy savings estimates. Other tools (e.g., DSMore)
are intended to provide program-level rather than site-
level estimates of energy savings.
Depending upon the level of detail desired, the tools
and methods described above have the ability to
produce detailed information about the clean energy
technology's patterns of operation. Building simulation
models, for example, produce detailed hourly load
patterns reflecting when an energy-efficient technology
reduces demand for a given building, application, and
climate zone. This information is needed to assess the
detailed impact on the utility system, specifically what
generation technology will be displaced or avoided
over the long term. Load shapes for particular technol-
ogies can also be acquired from third parties if building
simulations are not used.
Analysis of a renewables policy or program would
examine the costs and operation of eligible renew-
able resources and their interaction with the existing
(and planned future) generation system. This type of
analysis is often more complex, and therefore may
require a more sophisticated approach. A sophisticated
capacity planning and system dispatch model, for
example, would require information on the costs and
performance of renewables, as well as energy efficiency
options and their penetration potentials. Some of these
models have the ability to model energy efficiency
and renewable energy explicitly, reflecting potential
EE load shape impacts and penetration patterns, and
energy generation profiles for renewables. Others treat
these non-dispatchable and intermittent resources in
simpler ways.
Several sources are available to help predict the load
profile of different kinds of renewable energy and en-
ergy efficiency projects as listed below.
Performance data for renewable technologies are
available from the National Renewable Energy
Laboratory (NREL), as well as universities and oth-
er organizations that promote or conduct research
on the applications of renewable energy. For ex-
ample the Massachusetts Institute of Technology's
Analysis Group for Regional Energy Alternatives
and Laboratory For Energy and the Environment
conducted a 2004 report, Assessment of Emissions
Reductions from Photovoltaic Power Systems
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 39
-------
TABLE 2.2.6 EXAMPLES OF AVAILABLE TOOLS FOR ESTIMATING DIRECT ENERGY IMPACTS
Tool Name
Level of Analysis
Internet Based Methods
eCalc
New/retrofit
buildings
Renewable energy
sources (e.g., solar
heating, solar PV,
wind power)
Web-based calculator that enables users to design and
evaluate a wide range of clean energy projects for energy
savings and emissions reduction potential. In addition to
buildings and renewable energy sources, eCalc calculates
energy savings for municipal wastewater projects, traffic
lights, and street lighting projects.
h ttp://ecalc. tarn u. edu/
ENERGY
STARฎ Savings
Calculators
Energy efficiency
measures
Series of tools that calculate energy savings and cost
savings from ENERGY STAR-qualified equipment. Includes
commercial and residential appliances, heating and cooling,
lighting, office products, and other equipment.
h ttp://www. energys tar.
gov/'purchasing
ENERGY
STAR Roofing
Comparison
Calculator
Buildings
Estimates energy and cost savings from installing an ENERGY
STARฎ labeled roof product in a home or building.
http://www.roofcalc.com/
default.aspx
ENERGY STAR New buildings Helps planners, architects, and building owners set
Target Finder aggressive, realistic energy targets and rate a building
design's estimated energy use. Use the tool to determine:
Energy performance rating (1-100),
Energy reduction percentage (from an average building),
Source and site energy use intensity (kBTU/sf/yr),
Source and site total annual energy use (kBTU), and
Total annual energy cost.
Can use to evaluate potential energy savings of new/planned
buildings by building type for a clean energy policy (e.g., a
building code policy) and apply savings across the population.
h ttp://www. energys tar.
gov/targetfinder
ENERGYSTAR
Portfolio
Manager
Existing buildings
Portfolio of
buildings
Online, interactive tool that benchmarks the performance
of existing commercial buildings on a scale of 1-100 relative
to similar buildings. Tracks energy and water consumption
for building or portfolio of buildings and calculates energy
consumption and average energy intensity.
Can use to evaluate potential energy savings of existing
buildings by building type for a clean energy policy (e.g., a
building code policy) and apply savings across the population.
h ttps://www.energystar.
gov/benchmark
PVWatts
Grid-connected A solar technical analysis model available from NREL that
PV systems produces an estimate of monthly and annual photovoltaic
production (kWh) and cost savings. Users can select
geographic location and use either default system
parameters or specify parameters for their PV system. Data
can be used to accumulate project specific savings toward
renewable energy policy goals for solar-related technologies.
h ttp://rredc.nrel.gov/solar/
codes_algs/PVWATTS/
versionl/
Spreadsheet Based Methods
WindPro
Wind turbines
Wind farms
A Windows modular-based software suite for designing and
planning single wind turbines and wind farms.
EMD International,
WindPro: http://www.emd.
dk/WindPRO/lntroduction/
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 40
-------
TABLE 2.2.6 EXAMPLES OF AVAILABLE TOOLS FOR ESTIMATING DIRECT ENERGY IMPACTS fcontj
Tool Name
RETScreenฎ
Clean Energy
Project Analysis
Software
DSMore
Software Methods
fChart and PV-
fChart
eQuestฎ
ENERGY-10
DOE-2
Level of Analysis
Renewable energy
and energy
efficiency projects
DSM programs
Solar PV or solar
thermal systems
Buildings
Buildings
Buildings
Description
Use to evaluate the energy production and savings, costs,
emission reductions, financial viability, and risk for various
types of clean energy technologies, including renewable
energy, cogeneration, district energy, clean power, heating
and cooling technologies, and energy efficiency measures.
Designed to evaluate the costs, benefits, and risks of DSM
programs and services. Evaluates thousands of DSM scenarios
over a range of weather and market price conditions. While
requiring detailed input data, the model uses these data to
produce detailed outputs, including energy savings impacts
associated with the type of fuel that is being saved (gas or
electricity), and provides for expansive scenario analyses.
fChart Software produces both fChart and PV-fChart for
the design of solar thermal and photovoltaic systems,
respectively. Both programs provide estimates of
performance and economic evaluation of a specific design
using design methods based on monthly data.
Building simulation model for weather-dependent energy
efficiency measures. Energy savings can be applied across
the population.
Small commercial and residential building simulation models.
Can conduct a whole-building analysis, evaluating the energy
and cost savings that can be achieved by applying energy-
efficient strategies such as daylighting, passive solar heating,
and high-performance windows and lighting systems.
A building energy analysis computer program that predicts
the hourly energy use and energy cost of a building given
hourly weather information and a description of the building
and its HVAC equipment and utility rate structure.
http://www.retscreen.net/
ang/home.php
Integral Analytics:
http://www.
integralanalytics.com/
dsmore.php
h ttp://www. f chart, com/
index.shtml
http://www.doe2.com/
equest/
h ftp://www.nrel.gov/
buildings/energylO.html
http://www.doe2.com/
DOE2/index.html
(http://web. mit. edu/agrea/docs/MIT-LFEE_2004-
003a_ES.pdf). Another useful source is the Con-
necticut Energy Conservation Management Board
(http://www. ctsavesenergy. org/ecmb/index.php).
The California Database for Energy Efficient
Resources provides estimates of energy and peak
demand savings values, measure costs, and effec-
tive useful life of efficiency measures (http://www.
energy, ca.gov/deerf).
1 Some states or regions have technology produc-
tion profiles in their efficiency and renewable
energy potential studies (e.g., NYSERDA's
report, Energy Efficiency and Renewable Energy
Resource Development Potential in New York State,
2003, available at http://www.nyserda.org/sep/
EE&ERpotentialVolumel .pdf).
Load Impact Profile Data for energy efficiency
measures may be available for purchase from vari-
ous vendors, but typically is not publicly available
in any comprehensive manner.
1 Wind profiles can be obtained from a number of
sources, including the Department of Energy's
NEMS model (http://www.eia.doe.gov/oiaf/aeo/
overview/), NREL (www.nrel.gov), the American
Wind Energy Association (www.awea.org), and
several research organizations that have published
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 41
-------
information on wind resources in specific locations.
All data will likely require some extrapolation or
transposition for the intended use.
Considerations
When estimating the potential direct energy impacts,
states should consider the cost-effectiveness of the
measure or programs in the context of the avoided
costs12 of the utility system or region where they are
implemented. To evaluate cost-effectiveness, states can
conduct simple economic analyses such as project-level
discounted cash flow analysis. Using cash flow analysis,
the state develops estimates of the discounted cash flow
of alternative options reflecting any incentives available
under the program or policy, and simply compares
those with avoided costs (obtained from the PUC or
other entity, or estimated as discussed in Chapter 3)
in the region. For financial incentive-based programs,
measures that are less than the avoided cost (consider-
ing the incentive) could be expected to enter the mix.
For renewable mandates, technologies ranging from
least-to-most cost could be considered part of the po-
tential compliance set up to the minimum amount of
capacity required by the portfolio standard or goal.
It is important to remember, for this and more so-
phisticated methods, that there will be some degree of
non-compliance for certain mandated programs. For
example, building codes do not achieve 100 percent
compliance and enforcement is not complete. Calcula-
tions should factor non-compliance into the equation.
There are limits to this methodology. For example, the
revenue stream received by renewables will depend
on when they are operative (especially in competitive
markets). This method would miss the true distribu-
tion of costs that developers would face, and thus
would provide only a rough estimate of the financial
performance of these projects. It is important to note
that more sophisticated methods require this same
data for modeling the performance, economics, and
penetration of these technologies.
STEP 3.3: Determine Assumptions and
Review Available Data
Determining potential direct energy impacts attribut-
able to clean energy programs and policies requires
careful selection of assumptions based on state-
specific demographic and climatic conditions. Several
12 For more information about avoided costs, see Chapter 3, Assessing the
Electric System Benefits of Clean Energy.
assumptions should be considered when estimating the
prospective energy savings of a clean energy initiative.
These include:
Program period: What year does the program
start? End?
Program target: What sector or consumer type is
the focus of the program?
Anticipated compliance or penetration rate: How
many utilities will achieve the target or standard
called for? How many consumers will invest in new
equipment based on the initiative? How will this
rate change over the time period?
Annual degradation factor, how quickly will the
performance of the measure installed degrade or
become less efficient?
Transmission and distribution (T&D) loss: Is there
an increase or decrease in T&D losses that would
require adjustment of the energy savings estimate?
Adjustment factor: How should the estimate be
adjusted to factor in any inaccuracies in the calcula-
tion process?
Non-program effects: What portion of the savings
is due to factors outside of the initiative?
Funding and administration: What is the budget
for the program and how will it be administered?
What are the administrative costs? How much
will this reduce the amount of money available to
directly obtain energy savings?
Energy efficiency and renewable energy potential:
How do the savings projected compare to the po-
tential available? Are they realistic and consistent
with other relevant studies?
States can look to existing analyses to discover the as-
sumptions others have made while analyzing similar
programs. Multiple resources provide historical results
and projected EE and RE energy savings, including
those listed in Table 2.2.1. Other data sources include
the U.S. ENERGY STAR Program,13 the various utility
online audit services, and manufacturers and national
retailers. States can look to other state agencies (e.g.,
state energy and environmental offices) that may be
working on similar studies and have data on clean
energy estimates. Step 3.2 Choose Method to Estimate
' http://www.energystar.gov/
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 42
-------
Potential Direct Energy Impacts contains examples of
publicly available EE and RE data resources.
Additionally, states can assess available potential stud-
ies that support the clean energy policy decision. For
example, a potential study conducted for another state
may contain valuable information on the energy savings
associated with different clean energy programs, and
deemed savings databases from other states will include
energy savings for specific EE measures.14 Public service
commissions' Web sites usually post utility DSM filings
and Integrated Resource Plans, which contain details on
EE and RE plans with estimated energy savings.
In using data from other states or regions, it is impor-
tant to choose states that have similar climate and cus-
tomer characteristics. Even so, the assumptions about
operating characteristics of different clean energy tech-
nologies typically need to be adjusted for the specifics
of the state that is the focus of the study. For example,
for energy efficiency measures, adjustments for differ-
ences in weather are typically made, along with adjust-
ments for state-specific population characteristics.
STEP 3.4: Apply Model or Approach
In this step, states use the assumptions they develop
and apply the selected model or approach to the clean
energy initiative to estimate clean energy savings.
Examples of simple, bottom-up analyses of policy
options are presented below for appliance efficiency
standards, renewable portfolio standards, and lead by
example initiatives.
Air Conditioner Efficiency Standards
A state that is considering a new efficiency standard for
air conditioning could estimate energy savings based on
a variety of already-available data. The assessment could
use measure-specific energy savings from a deemed
savings database from another state (e.g., the California
Database of Energy Efficiency Resources), and adjust
the measure-specific savings to account for the weather
zones present in the state, especially for weather-
specific measures such as high-SEER air conditioning.
These adjustments might require the use of building
simulation models (e.g., eQuest; see Table 2.2.6) to
get reasonably accurate estimates of energy savings at
the site level. These site-level savings would ideally be
14 Deemed savings are validated estimates of energy savings associated with
specific energy efficiency measures that may be used in place of project-specific
measurement and verification.
generated for each housing type, air conditioning rating
level above federal standards, and weather zone. This
can create a large matrix of possible combinations.
Determining baseline market penetration of the higher
efficiency technology without conducting surveys
of HVAC dealers can be accomplished by reviewing
studies of market penetration rates from another
state (or states). These studies would need to be from
states that had not already adopted a higher efficiency
technology standard, and the results of the studies
would need to be adjusted for demographic differences
between the states.
Combined with some thoughtful analysis, these data
can help define the potential energy savings for the
proposed air conditioning measures without incurring
the time and expense of collecting all new data. Mak-
ing choices about which data to use and how to make
adjustments to those data involves inherent trade-offs
between the expected accuracy and the level of effort
expended. For example, using other states' existing
studies and applying basic adjustments to account
for different conditions would require less effort than
collecting region-specific data and developing savings
models for the local environment, but also would
be expected to yield a lesser degree of accuracy than
would the latter approach. Some analysis of the uncer-
tainty surrounding each key variable is recommended
in order to understand the relative accuracy of the
estimates obtained through these methods.
Renewable Portfolio Standard
In a similar manner, an estimate of the potential
energy savings associated with a renewable portfolio
standard (RPS) can use data from surrounding states
and/or those that have adopted similar rules regard-
ing the implementation of their RPS. For example, a
state might look at adoption rates for roof-mounted
solar photovoltaics in other states that have similar net
metering rules for solar systems and have established
incentives for installation that reward end-users and
developers in a similar manner financially15
Assumptions regarding the energy production of
the system, financial discount rate, and other factors
must be reviewed and projected in order to estimate
ls If the comparison state's financial incentives took the form of an upfront
rebate, and a future revenue stream based on RECs is assumed for the state
being analyzed, then a discounted cashflow analysis would be required to
analyze the net present value (NPV) of each approach to the project owner
and solar developer in order to compare the costs of the two approaches fairly.
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 43
-------
attractive rates of return that will stimulate the market
at the project level.
To extrapolate the project level analyses to the popula-
tion, factors including demographic data, the current
status of the solar industry in the state, and the current
economic climate are required to estimate a range of
savings that may be achieved through the policy over a
period of several years.
Lead by Example
To determine the energy savings from a lead by example
policy of reducing energy consumption in all state-
owned buildings 20 percent by 2020, a few basic steps
are required. The first is to gather the baseline data for
state-owned facilities, specifically their energy consump-
tion data for at least the past several years, along with
the square footage associated with each facility. These
data may take some time and effort to gather, as they do
not typically reside in one file or with one person.
Having the baseline data allows for summation of the
target kWh and therm reductions across all facilities.
If the policy will reduce energy consumption in exist-
ing buildings alone, calculating the savings number
is as simple as determining whether each facility will
achieve 20 percent savings, or the portfolio as a whole
will achieve a 20 percent reduction in annual con-
sumption. Either way, it is a straightforward exercise to
take 20 percent of the kWh and therms usage summed
for the base year. If the policy is to include new con-
struction as well, a determination of what the baseline
construction would have been for new state facilities in
the absence of the initiative, and an assessment of the
energy consumption associated with facilities built to
that evolving standard multiplied by the square footage
of planned additions, are needed.
To build a true bottom-up analysis of savings, though,
it is necessary to find where the 20 percent savings are
likely to come from. Individual building audits will pro-
vide the best data on where to achieve savings, and can
be summed by end-use, facility, and organization up to
the state level. But this process is relatively expensive and
time consuming, and a first-level screening might in-
volve benchmarking the facilities with national averages
and best-practice energy consumption per square foot.16
16 When benchmarking/utilities in this way, it is important to use bench-
marks specific to that building type. For example, a hospital has a very
different energy profile than does an office building, so only hospital-specific
benchmarks would be useful for benchmarking a hospital. See ENERGY
STAR's Portfolio Manager at http://www.energystar.gov/benchmark.
After initial screening, walk-through audits can be
used to confirm where to target the most cost-effective
initial investments. Most cost-effective energy efforts
start with lighting retrofits, as they are a proven energy
savings that can be easily achieved. Heating, ventilat-
ing, and air conditioning improvements or control
system upgrades will require a more detailed audit,
often take longer to complete, and require less modular
investments. Engineering algorithms or simulation
models are used to estimate the savings from HVAC
and other EE measures, and to estimate interactive
effects that may decrease the combined savings of indi-
vidual measures.
The level of detail desired may depend on the purpose
of the estimates. If, for example, agency budgets were
dependent upon their energy savings, a more detailed
analysis would provide better information about
specific technology performance and payback than a
screening-type of analysis. Regardless of the level of de-
tail, the state would sum up the measure and building
savings estimates across all facilities to assure that the
20 percent by 2020 statewide target can be met within
the budgets allocated.17
STEP 3.5: Evaluate Output
Once potential energy savings or generation impacts
are estimated, it is important to evaluate these results
to ensure that the numbers are reasonable and meet the
state's policy goals. If the state determines that the re-
sults are not realistic, it may need to review its assump-
tions and reapply the approach or model in an iterative
fashion to achieve reasonable energy savings estimates.
The resulting energy savings estimates can be compared
to a potential study, if available, to ensure that the policy
analysis does not overestimate the possible savings.
2.2.4 STEP 4.0: CREATE AN ALTERNATIVE
POLICY FORECAST
Once the direct energy impacts of clean energy are es-
timated, an alternative policy forecast must be created
that adjusts the BAU energy forecast developed under
Step 1 to reflect the clean energy policy or program.
In the case of efficiency, the energy savings estimates
would be subtracted from the BAU forecast to create a
17 Of course, other financing mechanisms for energy efficiency are available,
including bidding out the services to Energy Service Companies (ESCOs). This
chapter does not explore financing mechanisms, but focuses on energy savings
calculation methods and mentions the budget implications only as a consider-
ation for-policy makers.
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 44
-------
new forecast.18 For clean energy supply alternatives, the
policy forecast can be created with the sophisticated
supply forecasting models used to develop the original
BAU forecast (see Table 2.2.3). The assumptions in the
model would need to be adjusted to reflect the change
in renewable energy supply expected from the clean
energy initiative.
The impact estimates - and many of the same sophisti-
cated demand and supply models - can also be used to
assess impacts on the electric power system and project
what generation is likely to be displaced that otherwise
would have been in operation. This is discussed in
more detail in Chapter 3, Assessing the Electric System
Benefits of Clean Energy. In addition, the estimates can
also be used to determine environmental and economic
benefits as described in Chapters 4 and 5 respectively.
ISSUES AND CONSIDERATIONS
Incentives that are associated with the clean
energy policy can alter the energy savings esti-
mates (e.g., a renewable tax credit could increase
renewable energy production beyond RPS levels).
If historical trends do not reflect these incentives,
or non-economic based methods are used, states
should attempt to reflect the potential response to
these incentives.
Technologies change over time and can alter
energy savings estimates. This can alter the BAU
forecast and the potential for energy savings. BAU
forecasts and energy savings projections should be
reevaluated periodically (every one to two years).
This is particularly important under conditions of
rapid change.
Measurement and verification studies, which
estimate the actual energy savings of a clean energy
measure, can be used retrospectively to ensure that
an implemented clean energy program's perfor-
mance was reliably estimated and is meeting the
policy goals set out for the program.
As with all analyses, transparency increases
credibility. Be sure to document all sources and
assumptions.
18 Alternatively, two forecasts may be produced, with and without the clean
energy, and the difference would represent clean energy impacts. This meth-
odology would be more likely when using bottom-up economic-engineering
approaches.
2.3 CASE STUDIES
2.3.1 TEXAS BUILDING CODE
Impacts Assessed:
Electricity Savings
NOx Reductions
Clean Energy Program Description
The Texas Emissions Reduction Plan (TERP), initi-
ated by the Texas Legislature (Senate Bill 5) in 2001,
establishes voluntary financial incentive programs
and other assistance programs to improve air quality
[i.e., ozone formed from nitrogen oxides (NOx) and
volatile organic compounds (VOCs)] in the state. One
component of TERP recognizes the importance of
energy efficiency and renewable energy measures in
contributing to a comprehensive approach for meeting
federal air quality standards. Consequently, the legisla-
tion requires the Energy Systems Laboratory (ESL) at
the Texas Engineering Experiment Station of the Texas
A&M University System to submit an annual report
to the Texas Commission on Environmental Quality
estimating the historical and potential future energy
savings from energy building code adoption and, when
applicable, from more stringent local codes or above-
code performance ratings. The report also includes es-
timates of the potential NOx reductions resulting from
these energy savings. ESL has conducted this annual
analysis since 2002 and submits it in a report entitled
"Energy Efficiency/Renewable Energy Impact in the
Texas Emissions Reduction Plan." ESL also provides
assistance to building owners on measurement and
verification activities.
Method(s) Used
ESL determines the energy savings and resulting NO
emissions for new residential single- and multi-family
construction and for commercial office buildings in
Texas counties that have not attained federal air quality
standards. Its analysis is based on the energy efficiency
provisions of the IRC for single-family residences and
the IECC for all other residential and commercial
buildings. A brief summary of the approach for esti-
mating energy savings for both types of buildings is
provided below.
Residential Buildings. First, new construction activity
by county is determined. Next, annual and peak day
energy savings (in kWh) attributable to the building
code are modeled using a DOE-2 simulation that ESL
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 45
-------
developed for the TERR These estimates are then ap-
plied to National Association of Home Builders survey
data to determine the appropriate number of housing
types.
Commercial Buildings. The process to estimate energy
savings begins with estimating the number of buildings
and relative energy savings. The Dodge MarkeTrack
database provides construction start data and is used
to gather the square footage of new commercial con-
struction in Texas. These data are merged with energy
savings calculations published by the Pacific Northwest
National Laboratory (PNNL), along with the 1995
and 2003 Commercial Building Energy Consumption
database. The PNNL energy savings, which represent
buildings built to ASHRAE Standard 90.1-1989 versus
Standard 90.1-1999, are applied to the published square
feet of new construction.
After residential and commercial building savings
are estimated, these savings are projected to 2013 by
incorporating a variety of adjustment factors. These
factors include:
Annual degradation factor. This factor was used
to account for an assumed decrease in the perfor-
mance of the measures installed as the equipment
wears down and degrades. An annual degradation
factor of 5 percent was used for all the programs.
This value was taken from a study by Kats et al.
(1996).
T&D loss: This factor adjusts the reported savings
to account for the loss in energy resulting from
the transmission and distribution of the power
from the electricity producers to the electricity
consumers. For this calculation, the energy savings
reported at the consumer level were increased
by 7 percent to give credit for the actual power
produced that is lost in the transmission and
distribution system on its way to the customer. In
the case of electricity generated by wind, it was
assumed there was no net increase or decrease in
T&D losses, since wind energy is displacing power
produced by conventional power plants.
Initial discount factor: This factor was used to
discount the reported savings for any inaccuracies
in the assumptions and methods employed in the
calculation procedures. For single- and multi-
family programs, the discount factor was assumed
to be 20 percent.
Annual growth factor for single-family (3.25
percent), multi-family (1.54 percent), and for com-
mercial (3.25 percent) construction, derived from
recent U.S. Census data for Texas.
The state assumed that the same amount of electricity
savings from the code-compliant construction would
be achieved for each year after 2007 through 2013.
Results
The ESL 2008 annual report on the energy efficien-
cy and renewable energy impacts of the TERR sub-
mitted to the Texas Commission on Environmental
Quality in December 2008, describes prospective
energy savings resulting from implementing the
International Residential Code (IRC) and the In-
ternational Energy Conservation Code (IECC) in
residential and commercial buildings, respectively,
through 2020. According to the report, the cumu-
lative annual energy savings from code-compliant
residential and commercial construction were
estimated to be:
1,440,885 megawatt hours (MWh) of electricity
each year from 2001 through 2007, and
approximately 2.9 million MWh by 2013, account-
ing for 10 percent of the cumulative total electricity
savings under all energy efficiency and renewable
energy programs implemented under the TERP
between 2008 and 2013 (Texas A&M Energy Sys-
tems Laboratory, 2007).
ESL divided the actual and projected energy savings
into the different Power Control Authorities and, using
US EPAs eGRID emission factors, calculated the cumu-
lative annual NOx emission reduction values as follows:
1,014 tons-NO /year in 2007, and
X '
2,047 tons/year by 2013.
For More Information
Energy Efficiency/Renewable Energy Impact in The
Texas Emissions Reduction Plan (TERP). Volume
ISummary Report: Annual Report to the Texas
Commission on Environmental Quality. Janu-
ary 2007-December 2007. August 2008, Revised
December 2008. Energy Systems Laboratory, Texas
Engineering Experiment Station, Texas A&M
University System, http://esl.eslwin.tamu.edu/docs/
documents/tceq/ESL-TR-08-12-01%20tceq-report-
2007- Vol-I-FINAL.pdf
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 46
-------
Energy Efficiency/Renewable Energy Impact In The
Texas Emissions Reduction Plan (TERP). Prelimi-
nary Report: Integrated NOx Emissions Savings
From EE/RE Programs Statewide: Annual Report
to the Texas Commission on Environmental Qual-
ity January 2007-December 2007. August 2008.
http://esl. eslwin. tamu. edu/docs/documents/ESL-
TR-08-08-01.pdf
Development of a Web-based Emissions Reduc-
tion Calculator for Code-Compliant Commercial
Construction. Texas A&M Energy Systems
Laboratory. 2005. http://txspace.tamu.edu/
bitstream/handle/1969.1/5128/ESL-IC-05-10-34.
pdf?sequence=l.
Development of a Web-based Emissions Reduction
Calculator for Code-Compliant Single-Family and
Multi-Family Construction. Texas A&M Energy
Systems Laboratory. 2005. http://txspace.tamu.
edu/bitstream/handle/1969.1/5127/ESL-
IC-05-10-33.pdf?sequence=1.
2.3.2 VERMONT-ENERGY AND ENERGY
SAVINGS FORECASTING
Activities:
Energy forecasting
Energy savings forecasting
Background
The Vermont Department of Public Service (DPS) con-
ducts energy demand and energy efficiency program
savings forecasting as part of its long-term state energy
policy and planning process. This process includes:
The Comprehensive Energy Plan (CEP, required
under statute to be conducted every five years),
The 20-Year Electric Plan (also required every five
years), and
A variety of other state planning initiatives (Ver-
mont DPS, 2008).
The state uses the CEP as a tool to help manage the
transition from traditional energy fossil fuel to cleaner
energy supplies in order to benefit Vermont's economic
and environmental future. It provides a means for them
to show how energy demand and energy efficiency
program forecasts fit into the bigger planning picture.
TABLE 2.3.1 VERMONT PROJECTED
ENERGY DEMAND 2008-2010: WITH
AND WITHOUT NEW DSM
Without New
Year DSM (GWh)
2008 6,356
2009 6,324
2010 6,436
2011
2012
6,552
6,685
2013 6,821
2014
6,925
2015 6,941
2016 6,977
2017 7,042
2018
2019
2020
2021
7,123
7,205
7,293
7,381
2022 7,370
2023 7,440
2024 7,516
2025
2026
2027
2028
7,583
7,634
7,681
7,648
Total 148,933
AAGR
0.93%
With New DSM Energy Savings
(GWh) (GWh)
6,356 0
6,256 68
6,243 193
6,235
6,242
6,254
6,253
317
443
567
672
6,181 760
6,131 846
6,110 932
6,107
1,016
6,105 1,100
6,113 1,180
6,125
1,256
6,046 1,324
6,059
6,089
1,381
1,427
6,121 1,462
6,146 1,488
6,171 1,510
6,120
1,528
129,463 19,470
- 0.19%
AAGR=Average Annual Rate of Growth
Method(s) Used
For the 2008 study, the Vermont DPS began its
analysis by examining historical energy consumption
in Vermont across all sectors by selected fuel categories
between 1960 and 2005. It also uses the historical data
to compare energy demand in Vermont with demand
in New England and the United States from 1990
through 2004.
The process to forecast electricity and peak demand in
the state required several steps:
1. Determine fuel price projections and avoided costs
(i.e., the marginal energy supply costs that will be
avoided through savings in electricity, natural gas,
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 47
-------
and other fuels from a range of DSM programs.)
Consultants used DOE fuel price projections,
customized them to Vermont conditions, and de-
termined avoided costs using a screening tool that
contains load shapes for each measure and type of
program.19
2. Estimate the achievable, cost-effective potential for
electric energy and peak demand savings. The level of
efficiency potential in Vermont by DSM programs
was determined using the avoided cost estimates
from the first step along with various cost-effective-
ness tests (CDS, 2006).
3. Develop a 20-year forecast of electric energy use.
DPS hired consultants to develop a baseline projec-
tion of energy demand given current trends and
use patterns and a forecast of expected demand,
assuming implementation of the new DSM
measures, built up from estimates of energy use
by appliance type and end-use category by sector
(e.g., the number of refrigerators in the residential
sector) and the savings potential for each. Using
regression and trend analysis, Vermont ran one
20-year baseline forecast without new (projected)
DSM programs, and one case with assumed levels
of new DSM program activity.20
4. Develop a peak demand forecast. DPS also looked at
DSM savings using an econometric model base that
included historical DSM investments as an inde-
pendent variable. This method took a more conser-
vative approach than the regression analysis used
to project electric energy demand, in that it gives
equal weight to the past 20 years of DSM program
impacts and so may understate the credit deserved
by energy efficiency measures going forward.
Results
These historical data and the analysis show demand
for energy growing, driven by population growth,
economic development, larger homes, and increases
in vehicular travel. While overall energy demand ap-
peared to show more rapid growth in Vermont than for
the United States and New England, the reverse is true
19 The fuel cost and avoided cost assumptions were extensively reviewed by the
Avoided Energy Supply Component Study Group, composed of New England
utilities and PUCs.
20 The regression equation includes variables for personal income, price,
and trends to predict energy sales. The "with DSM" forecast was developed
by subtracting the DSM savings projections from the base case "without
DSM" forecast.
within the electricity sector, which has been the object
of intensive, formal energy efficiency program invest-
ments through Vermont's Energy Efficiency Utility.
In addition, Vermont faces a large supply gap if major
power contracts are not replaced, and the state projects
higher costs for new resources to replace them. In light
of this, Vermont committed itself to pursuing very ag-
gressive energy efficiency measures.
Based on the energy efficiency potential results
determined above, the DPS recommended DSM poli-
cies and a budget for programs. The Vermont Public
Service Board approved the budgets and the Efficiency
Utility established the specific programs (subject to
Public Service Board review).
The electricity forecasts projected that without new
DSM measures, electricity demand would grow an
average of 0.93 percent on an average annual basis
between 2008 and 2028. When new DSM measures are
implemented, the DPS anticipates that energy demand
will remain fairly flat, with a decline of 0.19 percent on
an average annual basis.
The Vermont DPS is currently developing a comprehen-
sive modeling approach using system dynamics (pos-
sibly relying on its older Energy 2020 model) to forecast
energy savings from its DSM programs that would,
ideally, better integrate the steps of its existing approach.
For More Information
Vermont Electric Energy Efficiency Potential Study,
Final Report. CDS Associates, Inc. May 10, 2006.
Prepared for the Vermont Department of Public
Service, http://www.state.vt.us/psb/document/
ElectricInitiatives/FinalReport-05-10-2006. doc.
Vermont Comprehensive Energy Plan 2009 and
Update to the 2005 Twenty-Year Electric Plan, Pub-
lic Review Draft. Vermont Department of Public
Service. May 2008. http://publicservice.vermont.
gov/planning/CEP%20%20WEB%20DRAFT%20
FINAL%206-4-08.pdf.
Vermont's Energy Forecasting Efforts. Vermont De-
partment of Public Service. June 19, 2008. http://
www.epa.gov/statelocalclimate/documents/pdf/
presentations_vt.pdf.
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 48
-------
California Energy Commission (CEC). 2005. Energy Demand Forecast Methods Report. http://www.energy.ca.gov/2005publications/
Companion Report to the California Energy Demand 2006-2016 Staff Energy Demand CEC-400-2005-036/CEC-400-2005-036.PDF
Forecast Report.
CEC. 2007. California Energy Demand 2008-2018, Staff Revised Forecast
h ttp://www. energy, ca. gov/20 07 p ublica tions/
CEC-200-2007-015/CEC-200-2007-015-SF2.PDF
Energy Information Administration (EIA). Commercial Buildings Energy Consumption http://www.eia.doe.gov/emeu/cbecs/
Survey.
EIA. Manufacturing Energy Consumption Survey.
EIA. Residential Energy Consumption Survey.
Elliott, R. Neal and Anna Monis Shipley. 2005. Impacts of Energy Efficiency and
Renewable Energy on Natural Gas Markets: Updated and Expanded Analysis.
American Council for an Energy-Efficient Economy. April.
Elliot, R. Neal and Maggie Eldridge. 2007. Role of Energy Efficiency and Onsite Renewables
in Meeting Energy and Environmental Needs in the Dallas/Fort Worth and Houston/
Galveston Metro Areas. American Council for an Energy-Efficient Economy. September.
Massachusetts. 2001. Energy Efficiency Activities 1999, A Report by the Division of
Energy Resources.
Myers, Stephen, James McMahon, and Michael McNeil. 2005. Realized and
Prospective Impacts of U.S. Energy Efficiency Standards for Residential Appliances:
2004 Update. Lawrence Berkeley National Laboratory, LBNL-56417. May.
New Jersey. Market Analysis and Baseline Studies.
Navigant Consulting Inc., Sustainable Energy Advantage LLC, and Boreal Renewable
Energy Development. 2004. New Jersey Renewable Energy Market Assessment
Final Report, August 2004, to Rutgers University Center for Energy, Economic and
Environmental Policy.
National Renewable Energy Laboratory (NREL). Energy Analysis - Slide Library.
NREL. Wind Research - Baseline Cost of Energy.
New York State Energy Research and Development Authprity (NYSERDA). 2005.
Executive Order No. Ill "Green and Clean" State Buildings and Vehicles. New York
State Energy Research and Development Authority.
Summit Blue Consulting. 2008. Assessment of the New Jersey Renewable Energy
Market, Volume 1 and II. Prepared for the New Jersey Board of Public Utilities. March.
Texas A&M Energy Systems Laboratory (ESL). 2005. Development of a Web-based
Emissions Reduction Calculator for Code-Compliant Commercial Construction.
Texas A&M ESL. 2005. Development of a Web-based Emissions Reduction Calculator
for Code-Compliant Single-Family and Multi-Family Construction.
http://www.eia.doe.gov/emeu/mecs/
http://www.eia.doe.gov/emeu/recs/
http://www.aceee.org/pubs/e052full.pdf
http://aceee.org/pubs/e078.pdf
http://www.mass.gov/Eoca/docs/doer/pub_
info/ee99-long.pdf
http://repositories.cdlib.org/lbnl/LBNL-56417/
http://www.njcleanenergy.com/main/public-
reports-and-library/market-analysis-protocols/
market-analysis-baseline-studies/renewable
http://policy.rutgers.edu/ceeep/
publications/2004/remareport.pdf
h ttp://www.nreigov/analysis/slide_library.h tml
http://www.nrel.gov/wind/coe.html
http://www.nyserda.org/programs/pdfs/
execorderlllfinalreportll-05.pdf
http://www.njcleanenergy.com/main/public-
reports-and-library/market-analysis-protocols/
market-analysis-baseline-studies/renewable
http://txspace.tamu.edu/bitstream/
handle/1969.1/5128/ESL-IC-05-10-34.
pdf?sequence=l
http://txspace.tamu.edu/bitstream/
handle/1969.1/5127/ESL-IC-05-10-33.
pdf?sequence=l
U.S. Environmental Protection Agency (EPA). ENERGY STAR. http://www.energystar.gov/
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 49
-------
Information Resources
URL Address
U.S. EPA. 2006. Clean Energy-Environment Guide to Action: Policies, Best Practices,
and Action Steps for States. April.
h ttp://www. epa.gov/sta telocalclima te/
resources/action-g uide.html
Vermont Department of Public Service. 2008a. Vermont's Energy Forecasting Efforts.
June 19.
h ttp://www. epa.gov/sta telocalclima te/
documen ts/pdf/presen ta tions_ vt.pdf
Vermont Department of Public Service. 2008b. Personal communications with Riley
Allen, Director of Planning, and Dave Lamont, Planning Engineer. 7/25/2008.
N/A
References URL Address
CDS Associates, Inc. 2006. Vermont Electric Energy Efficiency Potential Study, Final
Report. Prepared for the Vermont Department of Public Service. May 10.
Kats, G.H., Arthur H. Rosenfeld, and Scott A. McGaraghan. 1996. Energy Efficiency
as a Commodity: The Emergence of a Secondary Market for Efficiency Savings in
Commercial Buildings. European Council for an Energy Efficient Economy.
New Jersey. 2008. New Jersey Draft Energy Master Plan. April.
NYSERDA. 2005. New York Energy $MARTSM Program, Evaluation and Status Report for
the Year Ending December 2004. New York Public Service Commission and New York
State Energy Research and Development Authority. May.
NYSERDA. 2008. New York Energy $MARTSM Program, Evaluation and Status Report for
the Year Ending December 2007. New York Public Service Commission and New York
State Energy Research and Development Authority. March.
Texas A&M Energy Systems Laboratory (ESL). 2007. Energy Efficiency/Renewable
Energy Impact in the Texas Emissions Reduction Plan (TERP). Volume II- Technical
Report.
Vermont Department of Public Service. 2008. Vermont Comprehensive Energy Plan
2009 and Update to the 2005 Twenty-Year Electric Plan, Public Review Draft. May.
Wisconsin Office of Energy Independence. 2007. Wisconsin Energy Statistics.
h ftp://www.state.vt. us/psb/documen t/
Electriclnitiatives/FinalReport-05-10-2006.doc
http://www.eceee.org/conf erence_
proceedings/eceee/1997/Panel_2/p2_26/
Paper/
http://www.state.nj.us/emp/home/docs/pdf/
draftemp.pdf (draft plan) http://www.state.
nj.us/emp/ (a[[ related reports)
http://www.nyserda.org/Energy_lnformation/
SBC/sbcmay05summary.pdf
http://www.nyserda.org/pdfs/Combined%20
Report.pdf
http://txspace.tamu.edu/bitstream/
handle/1969.1/2077/ESL-TR-04-12-04.
pdf?sequence=l
http://www.publicservice.vt.gov/planning/
CEP%20%20WEB%20DRAFT%20FINAL%206-4-
08.pdf
http://energyindependence.wi.gov/docview.
asp?docid=116326flocid=160
CHAPTER 2 | Assessing the Multiple Benefits of Clean Energy 50
-------
CHAPTER THREE
Assessing the Electric System
Benefits of Clean Energy
Clean energy programs and policies can help states
achieve their goal of providing a less polluting, more
reliable and affordable electric system that addresses
multiple challenges, including:
Lowering energy costs for customers and utilities
alike, particularly during periods of peak electricity
demand;1
Improving the reliability of the electricity system
and averting blackouts at a lower cost;
Reducing the need for new construction of gener-
ating, transmission, and distribution capacity; and
Providing targeted reductions in load (i.e., the
amount of electric power or the amount of power
demanded by consumers at a given time) in grid-
congested areas, such as southwestern Connecticut
and San Francisco, California.
Many states are evaluating the electric system benefits
of clean energy. These benefits, as described above, go
beyond the direct energy savings and renewable energy
generation impacts discussed in Chapter 2, Assessing
the Potential Energy Impacts of Clean Energy Initiatives.
This chapter provides an overview of methods that can
be used to undertake broad assessments of the impacts
1 fust as energy efficiency program economics can be evaluated from a variety
of perspectives (total resource costs, program administration costs, ratepayer,
participant, and society) so can the benefits of clean energy programs. For each
perspective, the benefits of clean energy are defined differently. In this guide, we
are examining the equivalent of the total resource cost perspective, considering
benefits (and costs) to the participants and the utility. While other perspectives
including the utility costs are important, we focus on those perspectives most
important to policymakers and clean energy program administrators. For more
information about the different perspectives used to evaluate the economics of
programs, see Understanding Cost-Effectiveness of Energy Efficiency Programs:
Best Practices, Technical Methods, and Emerging Issues for Policy-Makers: A
Resource of the National Action Plan for Energy Efficiency, November 2008.
www.epa.gov/cleanenergy/documents/cost-effectiveness.pdf.
U
o
Q
CHAPTER ONE
Introduction
CHAPTER TWO
Potential Energy Impacts of Clean Energy
CHAPTER THREE
Electric System Benefits of Clean Energy
CHAPTER FOUR
Air Quality Benefits of Clean Energy
CHAPTER FIVE
Economic Benefits of Clean Energy
APPENDIX A
Catalogue of Clean Energy Case Studies
APPENDIX B
Tools and Models Referenced in Each Chapter
CHAPTER THREE CONTENTS
3.1 How Clean Energy Can Achieve Electric System
Benefits
53
3.1.1 The Structure of the U.S. Energy System 53
3.1.2 Primary and Secondary Benefits of Clean Energy....54
3.2 How States Can Estimate the Electric System Benefits
of Clean Energy 56
3.2.1 How to Estimate the Primary Electric System
Benefits of Clean Energy Resources 61
3.2.2 How to Estimate the Secondary Electric System
Benefits of Clean Energy Resources 77
3.3 Case Studies 83
3.3.1 California Utilities' Energy Efficiency Programs 83
3.3.2 Energy Efficiency and Distributed Generation in
Massachusetts 85
Information Resources 86
References 89
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 51
-------
STATES ARE QUANTIFYING THE ENERGY SYSTEM BENEFITS
OF CLEAN ENERGY POLICIES
Several states have quantified the energy system benefits from
their clean energy measures and determined that the measures
are providing multiple benefits, including avoiding the costs of
electricity generation, reducing peak demand, and improving
energy system reliability.
Georgia conducted an assessment of the benefits of achieving
energy efficiency improvements in the state and found it could
reduce demand for electricity by 3,339 GWh-12,547 GWh in
2010.
In addition to these energy savings, the analysis showed that the
improvements could benefit the overall electricity system and:
Avoid generation in Georgia of 1,207 GWh-4,749 GWh in
2010,
Reduce regional wholesale electricity cost by 0.5-3.9
percent by 2015, and
Lower peak demand by 1.7-6.1 percent by 2015 and
achieve a number of environmental and economic
benefits.
(Jensen and Lounsbury, 2005).
of clean energy on the overall electric system, including
effects on electricity generation, capacity, transmission,
distribution, power costs, and peak demand.
State legislatures, energy and environmental agencies,
regulators, utilities, and other stakeholders (e.g., rate-
payer advocates, environmental groups) can quantify
and compare the electric system benefits of clean
energy resources [e.g., energy efficiency, including
some demand response programs such as load control
programs, renewable energy, combined heat and
power (CHP), and clean distributed generation (DG)]
to traditional grid electricity. This information can
then be used in many planning and decision-making
contexts, including:
Developing state energy plans and establishing
clean energy goals;
Conducting resource planning (by PUCs or
utilities);
Developing demand-side management (DSM)
programs;
Conducting electric system planning, including
new resource additions (e.g., power plants), trans-
mission and distribution capacity, and intercon-
nection policies;
Planning and regulating air quality, water quality,
and land use;
Obtaining support for specific initiatives; and
Policy and program design.
Although quantifying electric system benefits can be
challengingparticularly when analyzing long-term
effects in a complex, interconnected electricity gridit
is important to consider these benefits when evaluating
clean energy resources. This chapter presents detailed
information about the energy system, specifically elec-
tricity benefits of clean energy, to help policy makers
understand how to identify and assess these benefits
based upon their needs and resources.
Section 3.1, How Clean Energy Can Achieve Electric
System Benefits, describes the energy system in the
United States and explains the multiple ways that
clean energy policies and programs can positively
affect the electric system and electricity markets,
thereby benefiting consumers, utilities, and society.
Section 3.2, How States Can Estimate the Electric
System Benefits of Clean Energy, presents an over-
view of the methods for estimating the primary
and secondary electric system benefits of different
types of clean energy resources.
Section 3.2.1, How to Estimate the Primary
Electric System Benefits of Clean Energy Re-
sources, describes the specific basic and sophis-
ticated modeling approaches and associated
tools that can be used to quantify a set of typi-
cally recognized (i.e., "primary") benefits.
Section 3.2.2, How to Estimate the Second-
ary Electric System Benefits of Clean Energy
Resources, describes approaches and tools for
estimating other electric system benefits (i.e.,
"secondary" benefits) that are less frequently
assessed and often more difficult to quantify.
Section 3.3, Case Studies, presents examples of how
two states, California and Massachusetts, are esti-
mating the electric system benefits of their clean
energy programs.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 52
-------
3.1 HOW CLEAN ENERGY CAN
ACHIEVE ELECTRIC SYSTEM BENEFITS
Energy is crucial to all aspects of the U.S. economy.
This section presents background information on how
the U.S. energy system is structured (see Section 3.1.1),
and describes the wide range of benefits that clean
energy can bring to the electricity component of this
system (see Section 3.1.2).
3.1.1 THE STRUCTURE OF THE U.S. ENERGY
SYSTEM
The energy system in the United States includes all
the steps, fuels, and technologies from the import or
extraction of energy resources, to their conversion to
useful forms, to their use in meeting end-use energy
demands (e.g., by the transportation, industrial,
residential, and commercial sectors). Components of
the energy supply system include transportation fuels,
electricity, and other forms of energy for use in homes,
manufacturing, and business. This chapter focuses on
several components of the larger electric system: elec-
tricity production, transmission, distribution, and the
markets by which electricity is bought and sold. These
components are hereinafter referred to together as the
electric system.
The North American electric system acts essentially
like four separate systems of supply and demand
because it is divided into four interconnected grids in
the continental United States and Canada: the Eastern,
Western, Quebec, and Electric Reliability Council of
Texas (ERGOT) Interconnections. These alternating
current (AC) power grids are depicted in Figure 3.1.1,
NERC Interconnections. Electricity can be imported or
exported relatively easily among the numerous power
control areas within each interconnection system.
However, for reliability purposes, the interconnections
have limited connections between them and are con-
nected by direct current (DC) lines.
Balancing the supply of and demand for electricity
in an economically efficient manner is complicated
by a number of factors. For example, the demand for
electricity varies significantly hour by hour, and cycli-
cally by time of day and season. Residential electricity
demand peaks in the morning and at night, when more
residents are at home and operating heating and air
conditioning units, washers, dryers, and other products
that use electricity. Commercial and industrial electric-
ity demand varies by type of company or industry, and
FIGURE 3.1.1 NERC INTERCONNECTIONS
NERC INTERCONNECTIONS
QUEBEC
INTERCONNECTION
WESTERN
INTERCONNECTION
EASTERN
INTERCONNECTION
ERCOT
NTERCONNECTION
Source: NERC, 2008.
thus may be considerably different from one location
to another.
Electricity supply is matched to demand using a port-
folio of production technologies. To meet the demand,
some power plants operate almost continuously, serv-
ing as baseload units (e.g., coal and nuclear plants are
examples of baseload units). Each baseload unit has
relatively high capital costs, but operational costs are
low. Also, startup and shutdown at these plants takes
time, is expensive, and causes additional wear on gen-
erating units. Other generation sources are operated
only during the times of highest demand, serving as
"peaking" units. The output of these generators rises
and falls throughout the day, responding to changing
electricity demand. Natural gas turbines are often used
for this purpose. These technologies are expensive to
run for long periods but can be started up and shut
down quickly. Because electricity must be generated at
the same time it is used, meeting peak demand and the
related price volatility are key issues.
The source of the electricity supply can also vary. A
group of system operators across the region decides
when, how, and in what order to dispatch electricity
from each power plant in response to the demand at
that moment and based on the cost or bid price. In reg-
ulated electricity markets, dispatch is based on "merit
order" or the variable costs of running the plants. In
restructured markets or wholesale capacity markets,
CHAPTER 3 | Assessing the Multiple Benefits of Clean Energy 53
-------
dispatch is based on the generator's bid price into the
market. Electricity from the power plants that are least
expensive to operate (i.e., the baseload plants) is dis-
patched first. The power plants that are most expensive
to operate (i.e., the peaking units) are dispatched last.
The merit order or bid stack is based on fuel costs and
plant efficiency, as well as other factors such as emis-
sions allowance prices.
Other conditions also affect electricity supply. Trans-
mission constraints (i.e., when transmission lines
become congested) can make it difficult to dispatch
electric generators located away from load centers and
move their power into areas of high demand, or may
require certain units to operate to improve system reli-
ability. Extreme weather events can decrease the ability
to import or export power from neighboring areas.
"Forced outages," when certain generators or transmis-
sion lines are temporarily unavailable, can also shift
dispatch to other generators. System operators must
keep all these issues in mind when dispatching power
plants. States can also take these issues into consider-
ation by using dispatch models or other approaches to
estimate which generators would likely reduce their
output and their emissions in response to the introduc-
tion of clean energy resources.
The electric power transmission system connects
power plants to consumers. Figure 3.1.2 depicts the
flow of power from the generating station, or power
plant, to the transformer and the transmission lines,
through the substation transformer (which reduces the
voltage) to the distribution lines, and finally, through
the pole transformer to the consumer's service box.
Electricity transmission is typically between the power
plant and a substation, and electricity distribution is the
delivery from the substation to consumers. Electricity
is usually transmitted through overhead transmission
and distribution lines, although sometimes under-
ground distribution lines are used in densely populated
areas. Overlapping lines are provided in the grid so
that power can be routed from any power plant to any
load center (e.g., populated areas), through a variety
of routes. Transmission companies conduct detailed
analyses to determine the maximum reliable capacity
of each line.
The process of generating, transmitting, and distribut-
ing electricity is quite complex and involves many
costs. Clean energy provides opportunities for states to
reduce many of those costs.
HOW ELECTRIC GENERATORS ARE DISPATCHED
The operation of electric systems is determined by a set of
physical constraints and economic objectives, through a
process referred to as "economic dispatch." The electric system
operator dispatches generating units (i.e., signals generators to
start or increase production) in economic merit orderthat is,
in order of increasing operating costs (starting with the lowest
costs adjusted for transmission losses), subject to reliability
considerations including transmission constraints. The highest-
cost unit dispatched at any point in time is said to be "on the
margin" and is known as the "marginal unit." For example,
high-cost combustion turbines and gas/oil peaking units are
on the margin for many hours of the week. During off-peak
times, plants with lower operating costs (e.g., combined cycle
gas turbines and coal-fired steam units) can be on the margin.
In some regions the cost used for dispatch is the variable cost
of running each plant (mainly fuel cost), but in others the
criterion for dispatch is a bid price submitted by the owners of
the generators.
3.1.2 PRIMARY AND SECONDARY BENEFITS
OF CLEAN ENERGY
Clean energy initiatives can result in numerous ben-
efits to the electric system, predominantly through
the avoidance of costs associated with generating,
transmitting, and distributing electricity. Clean energy
is often cheaper than or just as cost-effective as other
energy options, while delivering important electric
system, environmental, and/or economic benefits to
the state. For example, in California, energy efficiency
programs have cost the state 2
-------
FIGURE 3.1.2 FLOW OF ELECTRICITY FROM POWER PLANTS TO CONSUMERS
Color Key:
Blue: Transmission
Green: Distribution
Black: Generation
Transmission Lines
765, 500: 345,230, and 138 kV
Substation
Step-Down
Transformer
Generating Station
Generator Step
Up Transformer
Transmission
Customer
138kVor230kV
Subtransmission
Customer
26kV and 69kV
Primary Customer
13kVan
-------
emissions by allowing some units to shut down
and may delay or avoid the need for investment in
new generation to provide ancillary services. These
include stationary energy storage resources such as
batteries and pumped hydro storage. Other clean
energy resources, especially demand response
resourcessuch as controls on air conditioning or
water heater load control programscan free up
reserves that are needed to respond in the event
of a system outage. In some regions, clean energy
resources that operate during peak times reduce
the required level of operating resources.
Reduced wholesale market clearing prices. Clean en-
ergy policies and programs can lower the demand
for electricity or increase the supply of electricity,
causing wholesale markets to clear at lower prices.
This benefit can be dramatic during peak hours.
Increased reliability and power quality. An electric
grid is more reliable if the loads are lower, espe-
cially during peak hours and in areas where trans-
mission is constrained. Integration of clean energy
resources can increase the reliability of the electric-
ity system since power outages are less likely to
occur when the system is smaller and not strained;
more dispersed resources make the system less
vulnerable to outages. In addition, power quality
which is important for the operation of some
electrical equipmentcan be enhanced by some
forms of clean energy resources (e.g., fuel cells).
Avoided risks associated with long lead-time
investments. While clean energy resources certainly
have some risk (e.g., of underperformance of
energy efficiency or renewable energy measures),
these resources offer greater flexibility due to their
modular, segmented nature, and relatively quick
installation and disconnection time compared
with traditional resources. As a result, clean energy
options increase flexibility to deal with uncertainty
(relative to large, traditional fossil fuel resources)
by reducing dependence on conventional fuels
and allowing planners to be more responsive to
deviations from load forecasts. The size of the
potential for some clean energy options, such as
energy efficiency, is correlated with load, making
it especially responsive to changes in the planning
environment. In addition, reducing or delaying
the need for large utility investments for transmis-
sion or generation reduces both the need for large
amounts of financing and the chance of failed or
unnecessary investments.
Reduced risk from deferring investment in tradi-
tional, centralized resources until environmental and
climate change policies take shape. Clean energy
policies and programs may reduce the cost of
future compliance with air pollution control re-
quirements. In addition, clean energy policies and
programs may limit exposure to costs from any
future carbon regulations.
Improved fuel diversity and energy security.
Portfolios that rely heavily on a few energy
resources are highly affected by the unique risks
associated with any single fuel source (e.g., coal,
oil, gas). In contrast, the costs of some clean energy
resources are relatively unaffected by fossil fuel
prices and thus provide a hedge against fossil-fuel
price spikes. Other clean energy resources can be
affected by fossil fuel prices. For example, biomass
renewables may require fertilizer and/or process-
ing via technologies that use petroleum, natural
gas, and/or coal, and because wind provides
intermittent power that may not be available at
peak demand times, it can require backup peaking
units (e.g., natural gas turbines). Overall, however,
the greater the diversity in technology the less
likelihood of supply interruptions and reliability
problems. In addition, using diverse domestic
clean energy resources provides energy security by
reducing the vulnerability of the electric system to
attack and reducing dependence on foreign fuel
sources, such as imported petroleum, which may
yield political and economic benefits by protecting
consumers from supply shortages and price shocks
Table 3.1.1 summarizes the traditional costs of gen-
erating, transmitting, and distributing electricity, and
describes the primary and secondary clean energy
benefits associated with each type of cost.
3.2 HOW STATES CAN ESTIMATE
THE ELECTRIC SYSTEM BENEFITS OF
CLEAN ENERGY
The rigor with which states can or may want to analyze
the electric system benefits of clean energy depends
on the type of benefit being analyzed, the clean energy
proposal's status in the development and design pro-
cess, the level of investment under consideration, regu-
latory and system operator requirements, resources
(e.g., computers, staff) available for the analysis and,
for some benefits, the utility or region.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 56
-------
TABLE 3.1.1 ELECTRIC SYSTEM COSTS AND THE PRIMARY AND SECONDARY BENEFITS OF
CLEAN ENERGY
Traditional
Costs
PrimaryBenefits of
Clean Energy
Secondary Benefits of Clean Energy
Description of Benefit
Generation
Fuel
Variable operation
and maintenance
Emissions Allowances
Avoided costs
of electricity
generation or
wholesale electricity
purchases.
Reduced risk from investment in
traditional, centralized resources
before environmental and climate
change policies take shape.
Improved fuel and energy security.
Avoided ancillary services.
Reductions in wholesale market
clearing prices.
Increased reliability and power
quality.
Avoided risks associated with long
lead-time investments (e.g., risk of
overbuilding the electric system).
Clean energy policies and
programs can displace
traditional electric energy
generation.
Clean energy policies and
programs can lower the
demand for electricity
or increase the supply of
electricity, causing wholesale
markets to clear at lower
prices.
Section
3.2.la
Capital and operating
costs of upgrades
Fixed operation and
maintenance
New construction to
increase capacity
Transmission & Distribution
Avoided costs
of power plant
capacity.
Clean energy policies and
programs can delay or avoid
the need to build or upgrade
power plants.
Capital and operating
costs of maintenance
Upgrades
New construction
Deferred or
avoided costs of
transmission &
distribution (T&D)
capacity.
Increased reliability and power
quality.
Clean energy policies and
programs that are located close
to where energy is consumed
can delay or avoid the need to
build or upgrade T&D systems.
3.2.1b
3.2.1c
Energy losses
Avoided electric
loss in T&D lines.
Clean energy policies and
programs that avoid energy
consumption also avoid losses
associated with transmission
and distribution.
3.2.Id
A range of basic and sophisticated methods is available
to allow analysts to estimate how the electric system
will be affected by clean energy measures, including
when and where electricity generation may be offset.
Basic methods typically include spreadsheet-based
analyses or the adaptation of existing studies or infor-
mation. Sophisticated methods typically use dynamic
electric system models that (a) predict the response of
energy generation to actions that influence the level of
clean energy resources and (b) calculate the resulting
effects. These two approaches are not mutually exclu-
sive, but may be used in a complementary way. Table
3.2.1 describes the advantages and disadvantages of
each method and when they are appropriate to use.
SELECTING BENEFITS TO EVALUATE
Some states may not be interested in estimating all
types of electric system benefits, or states maybe
considering programs that deliver benefits in only
some areas. It is generally common practice to evaluate
CHAPTER 3 | Assessing the Multiple Benefits of Clean Energy 57
-------
TABLE 3.2.1 ADVANTAGES AND DISADVANTAGES OF BASIC VS. SOPHISTICATED METHODS OF
ESTIMATING ELECTRIC SYSTEM BENEFITS
Advantages
Basic Estimation
Relatively low cost.
Requires minimal input data and time.
Sophisticated Simulation
Robust representation of electric system
dispatch and, in some cases, capacity
expansion.
Provides high level of analytic rigor and
detailed results.
May be available from utility resource
planners.
May allow sensitivities to a wide range of
assumptions.
Disadvantages
Less robust.
Provides approximate estimates.
Time- and resource-intensive.
Relatively high cost.
Requires significant input data.
Complex.
Not transparent in stakeholder process.
When to Use
For preliminary studies.
When time and/or budget are limited.
When limited data resources are available.
When a high degree of precision and
analytic rigor is required.
i When sufficient data resources are
available.
all the primary benefits for clean energy projects or
programs. For secondary benefits, however, the need
for detailed estimation can vary depending on several
factors, including:
The type of clean energy resource being considered,
Regulatory or system operator study requirements,
Available resources (e.g., computers, staff, and data),
and
Whether certain needs or deficiencies have been
identified for the existing electric system.
For example, suppose a state is considering demand
response resources such as direct load control (i.e.,
programs that enable electric providers to reduce the
demand of consumer sites at peak times, sometimes by
directly curtailing major energy-intensive equipment
such as air conditioners and water heaters). For these
types of measures, it is increasingly common to con-
sider wholesale market price effects because the benefit
to consumers from price reductions during peak hours
can be substantial. On the other hand, if a state energy
efficiency policy is expected to produce significant
savings only during off-peak hours or seasons, which
would result in a smaller impact on the wholesale mar-
ket, it may not be worthwhile to estimate the wholesale
market price effects. Similarly, quantification of
ancillary service benefits can be difficult in areas
without regional transmission organizations (RTOs)
that routinely report market prices, even if the clean
energy resource has the capability of delivering these
ancillary service benefits. In this case, analysts may
decide to devote their limited staff and computing
power to quantifying benefits that are likely to yield
the most reliable and meaningful results, and address
other benefits qualitatively.
There are a number of considerations in selecting
which benefits to estimate. As indicated earlier, prima-
ry electric system benefits tend to be easier to quantify
and the methods to quantify them tend to be mature.
The methods to evaluate the secondary electric system
benefits are more limited and can be subject to debate.
Tables 3.2.2 and 3.2.3 outline some of the factors that
states can consider when deciding which electric sys-
tem benefits to analyze, including available methods
and examples, advantages, disadvantages, and purpose
of analysis. Section 3.2.1, How to Estimate the Primary
Electric System Benefits of Clean Energy Resources, and
Section 3.2.2, How to Estimate the Secondary Electric
System Benefits of Clean Energy Resources, review each
type of benefit and explain the approaches generally
used to analyze each benefit.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 58
-------
TABLE 3.2.2 PRIMARY ELECTRIC SYSTEM BENEFITS FROM CLEAN ENERGY MEASURES
Applicable Clean
Energy Resources
Considerations for Determining
Whether to Analyze
Who Usually Conducts Analysis?
When is Analysis
Usually Conducted or
Made Available?
BENEFIT: Avoided electricity generation or wholesale electricity purchases
All resources.
Resources that
operate during peak
hours.
Traditionally analyzed in cost-benefit
analysis.
Widely accepted methods.
Data generally available but expensive.
Models available but are complex, not
transparent, and are often expensive
to use.
Many assumptions about technology,
costs, and operation needed.
Long term fuel price forecasts must be
purchased or developed.
Utilities conduct in-depth
modeling.
PUCs and other stakeholders review
utility's results and/or conduct own
analysis.
RTO/ISO and the Independent
Market Monitor.
US EIA and private consultancies
provide electric dispatch and
capacity expansion forecasts.
Resource planning and
released regulatory
proceedings.
Area-specific DSM
program development.
RTO/ISO avoided cost
estimates may be
published on regular
schedules.
BENEFIT: Avoided power plant capacity additions
All resources.
Resources that
operate during peak
hours.
Traditionally analyzed in cost-benefit
analysis.
Generally accepted methods for both
estimation and simulation.
Some assumptions about technology,
costs and operation needed.
Data generally available.
Utilities conduct in-depth
modeling.
PUCs and other stakeholders review
utility's results and/or conduct own
analysis.
In some regions, RTO/ISO publishes
capacity clearing prices.
Resource planning and
proceedings.
Area-specific DSM
program development.
RTO/ISO avoided cost
estimates may be
published on regular
schedules.
BENEFIT: Deferred or avoided T&D capacity
Resources that
are close to load,
especially those that
operate during peak
hours.
Traditionally analyzed in cost-benefit
analysis.
Load flow forecast availability.
Unit cost of T&D upgrades can be
estimated but may be controversial.
T&D capacity savings reasonably
practical, but site-specific savings
difficult to generalize.
Utilities conduct in-depth
modeling.
PUCs and other stakeholders review
utility's results and/or conduct own
analysis.
. RTO/ISO.
T&D build planning.
Area-specific DSM
program development.
RTO/ISO costs
estimates may be
published on regular
schedules.
BENEFIT: Avoided energy loss during T&D
Resources that
are close to load,
especially those that
operate during peak
hours .
Traditionally analyzed in cost-benefit
analysis.
Straightforward; easy to estimate
once avoided energy has been
calculated
Loss factor for peak savings may
need to be estimated.
Utilities collect loss data regularly
and may conduct in-depth
modeling.
PUCs and other stakeholders review
utility's results and/or conduct own
analysis.
and
Resource planning and
proceedings.
Area-specific DSM
program development.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 59
-------
TABLE 3.2.3 SECONDARY ELECTRIC SYSTEM BENEFITS FROM CLEAN ENERGY MEASURES
Applicable Clean
Energy Resources
Considerations for Determining
Whether to Analyze
Who Usually Conducts Analysis?
When is Analysis
Usually Conducted?
BENEFIT: Avoided Ancillary Services
Resources that can
start during blackout,
ramp up quickly,
or provide reactive
power.
Resources closer to
loads.
Usually smaller benefits than
traditionally analyzed benefits .
Market price data available for some
services in some markets (e.g., PJM).
Ancillary service savings from clean
resources often site-specific and
difficult to estimate.
Separating ancillary service value
from capacity value in long run
analysis may be difficult.
Utilities conduct in-depth modeling.
PUCs and other stakeholders review
utility's results and/or conduct own
analysis.
Resource planning and
proceedings.
Area-specific DSM
program development.
BENEFIT: Wholesale Market Price Effects
All clean resources .
Resources that
operate during peak
hours.
Benefits depend on market/pricing
structure and peaking resources and
forecasted reserve margins.
Actual market price data generally
available.
Studies to estimate benefits may be
complex.
ISOs and utilities conduct in-depth
modeling.
PUCs, other stakeholders review
utility's results and/or conduct own
analysis.
Resource planning and
proceedings.
Area-specific DSM
program development.
Policy studies.
BENEFIT: Increased reliability and power quality
Distributed resources.
Resources close to
load or with high
power quality.
All resources that
operate as baseload
units.
All load reducing
resources that
increase surplus
generating and T&D
capacity in region.
Historical reliability data often
available.
Historical power quality data rare.
Studies for converting to dollar value
complex and controversial.
Benefits are especially valuable
for manufacturing processes that
are sensitive to power quality or
regions where reliability is significant
concern.
Utilities conduct in-depth modeling
PUCs and other stakeholders review
utility's results and/or conduct own
analysis.
Usually ad hoc studies.
Distributed resources
with short lead times.
Resources close to
load
All clean resources.
Historical load and load variability
data often available.
Modeling varies from simple to
complex.
Utilities conduct in-depth modeling.
PUCs and other stakeholders review
utility's results and/or conduct own
analysis.
Policy and risk management
analysts.
Resource planning and
regulatory review of
planning.
Policy studies.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 60
-------
TABLE 3.2.3 SECONDARY ELECTRIC SYSTEM BENEFITS FROM CLEAN ENERGY MEASURES fcontj
Applicable Clean
Energy Resources
Considerations for Determining
Whether to Analyze
Who Usually Conducts Analysis?
When is Analysis
Usually Conducted?
environmental and climate change policies are implemented)
All clean energy
resources.
Modeling varies from simple to
complex.
Studies to estimate benefits may be
complex.
Regulatory uncertainty adds to
complexity of analysis.
Fuel and technology diversification
All clean energy
resources.
i Diversity metrics computable from
generally available data
Portfolio analysis of costs vs. risks
adds complexity.
i Must consider existing supply
resources, not just incremental new
resources.
Policy and risk management
analysts.
Resource planning and
regulatory review of
planning.
Policy studies.
States.
PUCs.
Utilities.
State energy plans.
Resource planning.
:
3.2.1 HOW TO ESTIMATE THE PRIMARY
ELECTRIC SYSTEM BENEFITS OF CLEAN
ENERGY RESOURCES
Implementing clean energy policies and programs
results in reduced demand for electricity. As described
earlier, the primary electric system benefits resulting
from this reduced demand include:
Avoided cost of energy generation or wholesale
energy purchases,
Avoided cost of power plant capacity,
Deferred or avoided T&D capacity costs, and
Avoided energy loss during T&D.
States can compare different electric resources, includ-
ing clean energy resources such as energy efficiency,
renewable energy, clean distributed generation, or
combined heat and power, by examining the net
present value of the revenue requirements over the
life of the resource. This enables comparison of
various options on an equal basis, combining capital
investmentsaccounting for carrying costs over the
book life of the investmentwith the discounted
value of their annual fuel and operating costs over the
investment's operating life. For example, installing
high-efficiency transformers in a new substation
can be more expensive than standard equipment in
terms of up-front costs, but will waste less electricity
over time, thereby reducing variable operating and
maintenance costs. Likewise, replacing a chiller in a
food-processing factory with a more efficient unit in-
curs a higher capital cost up-front, but reduces annual
electricity costs for the customer.3 The basic concept
is to compare the net impact on the cost of power over
the lifetimes of each alternative that is technically
capable of meeting the need. The alternative with the
smallest net impact is typically the preferred choice, all
other things being equal.
As indicated above, methods to quantify primary elec-
tric system benefits are mature and states can choose
from a range of basic and sophisticated methods as
described below.
3 Some states have competition in retail electricity service, others do not,
and some are in a transitional state. These examples apply to both traditional,
vertically integrated utilities and to distribution-only utilities. However, the
existence of retail competition changes some of the details in important ways.
One such difference is that under retail competition, a portion of the cost
savings from lowering electric consumption accrues to the distribution utility
(e.g., reduced need to expand T&D lines) and a portion becomes a reduction
in the revenues of competitive wholesale generators. The policy implications
of that split need to be considered, but the important point is that the entire
savings accrues to the retail customers and to society as a whole.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 61
-------
Basic Methods
Basic methods span a broad range of possibilities, but
generally rely on relatively simple relationships and
analytic structures. Many are conceptually similar to
sophisticated methods, but they use simplifying as-
sumptions (proxy plants, system averages) rather than
using detailed models to develop the impacts or pa-
rameters to estimate impacts (e.g., emissions factors).
For example, in order to estimate impacts of a clean
energy resource, the goal is to match impacts (in terms
of reduced demand for electricity) to the generation
resource that will be displaced. However, instead of
running a dispatch model to make these estimates,
simple proxiesfor generating units displaced, or
emissions rates at the time of displacementare used
instead. A dispatch model would identify specifically
those units on the margin in each time period, but with
a basic method it may be sufficient to pair impacts (i.e.,
changes in generation requirements due to energy ef-
ficiency or other clean energy resources) to the general
type of unit expected to be on the margin. For example,
for all impacts during the peak period, a natural-gas-
fired combustion turbine could be used to estimate
impacts. During baseload periods, a coal plant could
be used; while in shoulder periods an oil/gas steam
might be used. The details would depend on the system
being analyzed.
Estimation methods can be used for preliminary
assessments or screening exercises, such as compar-
ing the cost of a clean energy option with a previous
projection of avoided costs or the cost of a proxy plant.
Proxy plant assessments are typically done using cost
assumptions for the expected next addition; for exam-
ple, a natural gas combined cycle plant. Although they
are less robust than modeling methods, basic methods
require less data, time, and resources, so they can be
useful when time, budget, and data are limited.
Sophisticated Methods
State-of-the-art power sector models for simulating
and projecting power plant operations and costs (or
T&D system adequacy) represent one type of sophis-
ticated model. The sophisticated models have more
complex structures and interactions than the basic
approaches, and are designed to capture fundamental
behavior of the sector using engineering-economic
relationships or econometric approaches. They require
additional input assumptions compared with basic
methods, but add the ability to evaluate how the
operations and capacity needs of the existing electric
grid will change with the adoption of a clean energy
resource, based on engineering and economic funda-
mentals. Some models can predict energy prices, emis-
sions, and other market conditions as well.
These models are complex to set up and can be costly.
Developing a detailed representation of the electric
system can involve many individual input assump-
tions, and it is important to validate, benchmark, or
calibrate complex models against actual data. Access
to confidential system data can also pose a challenge
to conducting rigorous avoided cost analysis. How-
ever, in many cases datasets already exist for regional
and utility planning analyses. Furthermore, existing
sector models have the benefit of being well under-
stood and mature.
While developing a full input data set for a dispatch
simulation model can be a daunting task, it can provide
a higher level of analytic rigor than basic estimation
methods, which simplify complex systems and can
result in errors in estimated costs. It is important to con-
sider whether existing utility models can be relied on
and are acceptable to stakeholders in a stakeholder pro-
cess. If they can be relied on, the incremental work of
estimating clean energy benefits will be greatly reduced.
Simulations of clean energy programs using sophisti-
cated models can be done on an individual basis (e.g.,
modeling the impact of wind turbines) or the analysis
can be used to assess multiple clean energy strategies.
A single analysis of an affected system can provide a
basis for analyses of a large number of clean energy
programs simultaneously. For example, a sophisticated
model may have the ability to assess the impact of an
energy efficiency program and a renewable portfolio
standard, capturing any interactions between the two.
One of the benefits of more sophisticated approaches is
their ability to capture these kinds of interactions.
The remainder of this section provides details about
the methods available to assess the four primary elec-
tric systems benefits of clean energy.
3.2.1.a Avoided Costs of Electricity Generation
or Wholesale Electricity Purchases
New clean energy resources (on the demand and sup-
ply side) avoid electricity and capacity costs in both
the short run (e.g., three years or less) and in the long
run (e.g., typically five to 20 years). In the short run,
avoided costs consist of avoided fuel, variable operation
and maintenance (O&M), and emissions allowances
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 62
-------
that can be saved at those generating units that would
operate less frequently as a result of new clean energy
resource additions. Methods to estimate these short-
run avoided costs are described in this section.
In the long run, however, avoided costs consist largely
of the capital and operating costs associated with new
generation capacity and T&D capacity that are dis-
placed or deferred by clean energy resources.4 Meth-
ods to estimate these long-run costs are described in
Section 3.2.1.b, Avoided Costs of Power Plant Capacity,
and Section 3.2.1.C, Avoided Transmission and Distri-
bution Capacity.
Key Considerations
A number of challenges arise when calculating short-
and long-run avoided costs. Avoided cost estimates
generally depend upon the comparison of two cases:
A baseline or reference case without the new re-
source, and
A case with the new resource, which in the case of
a demand-side resource includes a reduction in the
load or load decrement.
4 Sometimes the short-run and long-run effects of dean energy measures are
referred to as "operatingmargin"and "build margin',' respectively (Biewald, 2005).
Short-run avoided costs of electricity generation are the
operating costs of marginal units. Operating costs include
fuel, variable O&M, and marginal emission costs. In a
competitive market, wholesale energy prices will reflect
the generator's actual costs for operating marginal units
in the bids they submit.
Consequently, both cases involve projections of future
conditions and are subject to many uncertainties that
influence electricity markets (e.g., fuel prices, construc-
tion costs, environmental regulations, and market
responsiveness to prices). Since avoided costs are cal-
culated as the difference between these two cases, they
can be very sensitive to the underlying assumptions for
either or both cases. This uncertainty is characteristic
of long-run avoided cost calculations which require
projections far out into an uncertain future. Therefore,
states may want to consider performing sensitivity or
scenario analyses on both the underlying base case
(e.g., on demand growth, fuel prices) and on the key
drivers of the case with the new resources (e.g., on the
cost or timing of new resources) to gauge the potential
range of results.
TABLE 3.2.4 COMPARISON OF BASIC AND SOPHISTICATED APPROACHES FOR QUANTIFYING AVOIDED
COST OF ELECTRICITY GENERATION OR WHOLESALE ELECTRICITY PURCHASES
Proxy unit
Futures prices
Previously estimated
cost projections
Simple.
May already be
available.
Sophisticated Method (Dispatch Modeling)
ProMod
Market Analytics
MAPS
. IPM
Robust representation
of electrical system
dispatch.
Combines energy & capacity.
Not always relevant to a given policy
if timing or costs are different.
Limited horizon (futures).
May miss interactive effects (fuel
and emissions markets) and leakage
effects for significant clean energy
investments over time.
When time, budget and data are
limited.
Rough estimates.
Preliminary assessment.
Overview-type policy assessment.
Cost.
Data- and time-intensive.
Not transparent.
When clean energy resource use
will change system operations (e.g.,
clean energy resources change the
marginal generating resource in a
large number of hours).
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 63
-------
Methods for Estimating Short-Run Avoided Costs
of Electricity Generation or Wholesale Electricity
Purchases
Two types of methods for quantifying short-run avoid-
ed costs of electricity generation or wholesale electric-
ity purchasesbasic and sophisticatedare described
below. Both have advantages and limitations that are
dictated by individual circumstances (see Table 3.2.4),
and involve these steps as presented in Figure 3.2.1.
1. Estimate clean energy operating characteristics.
Using the total energy impacts estimates (as
described in Chapter 2), estimate the load impact
or energy generation profile of the clean energy
measurean estimate of when the energy would
be availableeither on an hourly basis, or some
other more aggregate time scale.
2. Identify the marginal units to be displaced. Identify
the generation resources that would be displaced
as a result of the clean energy resource, either due
to reduced demand or increased supply of clean
energy.
3. Identify the characteristics of the marginal units
displaced. This specifically includes the avoided
energy costs (and as described later, avoided
emissions).
4. Map the energy impacts to the displaced unit in-
formation. This is done to calculate the short-run
avoided costs of electricity generation. For basic
methods, the estimated energy impacts (reduc-
tion in load or energy supplied) are mapped to
the displaced energy information. For example, if
hourly impacts are estimated, hourly kWh savings
are multiplied by hourly avoided costs estimates.
The summation of these hourly values represents
the impact of the clean energy resource on costs.
For sophisticated methods, this calculation may be
a direct output of the modeling exercise.
The various approaches are described further below.
Basic Methods for Estimating Short-Run Avoided
Costs
Short-run avoided costs of energy generation can be
estimated using simplified methods, such as spread-
sheet analysis of market prices, marginal cost data, or
inspection of regional dispatch information (i.e., fuel
mix and capacity factor by fuel type). Non-modeling
estimation methods, such as using a previously esti-
mated avoided cost projection, maybe more appropri-
ate when time, budget, and access to data are limited,
but they result in an approximation of the costs of
avoided energy generation. Consequently, it is impor-
tant for analysts to consider whether the estimation
method is an acceptable representation of the actual
system. For example, already-available avoided costs
may be out of date or may not match the timing of the
impacts of the clean energy resource being considered.
The general steps involved in conducting these meth-
ods are described in more detail below.
Step 1: Estimate clean energy operating characteristics.
The first part of estimating avoided costs of clean en-
ergy is to estimate the amount of energy (in kWh) the
clean energy measure is expected to generate or save
over the course of a year and its lifetime. Methods for
estimating this were described in Chapter 2.
In addition to estimating annual impacts, it maybe de-
sirable to estimate the timing of impacts within a year,
either hourly or on some less frequent interval. Clean
energy resources that reduce generation requirements
at the time of peak, when combustion turbines may be
operating, will differ from those that affect the system
during periods of low demand when oil/gas steam
plants or coal plants may be operating.
In the case of energy efficiency measures, load impact
profiles describe the hourly changes in end-use de-
mand resulting from the program or measure. In the
case of energy resources, the generation profiles (for
wind or PV, for example) are required. The time period
can range from 8,760 hourly intervals to two or three
intervals, such as peak, off-peak, and shoulder periods.
Similarly, a wind turbine can be expected to produce
differing quantities of electricity across the day and
year. These data are used to identify more precisely
what specific generation or generation types are dis-
placed by the clean energy resources.
Several sources are available to help predict the load
profiles of different kinds of renewable energy and
energy efficiency projects:
Performance data for renewable technologies are
available from the National Renewable Energy
Laboratory (NREL), as well as universities and oth-
er organizations that promote or conduct research
on the applications of renewable energy. For ex-
ample, the Massachusetts Institute of Technology's
Analysis Group for Regional Energy Alternatives
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 64
-------
FIGURE 3.2.1 STEPS FOR ESTIMATING
AVOIDED COST
STEPl
Estimate Clean Energy Operating Characteristics
STEP 2
Identify the marginal units to be displaced
I
bee
A
STEPS
Identify the operating costs of marginal units to be displaced
1
STEP 4
Calculate the short-run avoided costs of electricity generation
and Laboratory For Energy and the Environment
published a report in 2004 entitled Assessment of
Emissions Reductions from Photovoltaic Power
Systems (http://web. mit. edu/agrea/docs/MIT-
LFEE_2004-003a_ES.pdf). Another useful source
is the Connecticut Energy Conservation Manage-
ment Board (http://www.ctsavesenergy.org/ecmb/
index.php).
' The California Database for Energy Efficient Re-
sources (DEER) provides estimates of energy and
peak demand savings values, measure costs, and
effective useful life of efficiency measures (http://
www. energy, ca.gov/deer/).
' Some states or regions have technology produc-
tion profiles in their efficiency and renewable
energy potential studies (e.g., NYSERDAs report,
Energy Efficiency and Renewable Energy Re-
source Development Potential in New York State,
2003, available at http://www.nyserda.org/sep/
EE&ERpotentialVolumel .pdf).
Load impact profile data for energy efficiency mea-
sures may be available for purchase from various
vendors, but typically is not publicly available in
any comprehensive manner.
1 Wind profiles can be obtained from a number of
sources, including the Department of Energy's
NEMS model (http://www.eia.doe.gov/oiaf/aeo/
overview/), NREL (www.nrel.gov), the American
Wind Energy Association (www.awea.org), and
several research organizations that have published
information on wind resources in specific locations.
All data will likely require some extrapolation or
transposition for the intended use.
In the absence of specific data on the load impact or
energy profile of the clean energy resource, analysts
will need to use their judgment to assess the timing of
that resources impacts.
Step 2: Identify the marginal units to be displaced.
The next step is to identify the units and their associ-
ated costs that are likely to be displaced by the clean
energy resources. While this section discusses the
process of estimating avoided cost benefits, these same
methods support the estimation of emissions benefits
of clean energy.
In each hour, electric generating resources are dis-
patched from least to most expensive, on a variable cost
basis, until demand is satisfied. There are a host of com-
plexities involved in dispatching the generating system,
including generator start-up and shut-down operating
constraints and costs, and transmission and reliability
considerations, among other factors. However, in
concept, the unit that is displaced is the last unit to be
dispatched. Estimating the benefits of clean energy
resources requires identifying this "marginal" unit and
its avoided costs. Because reported or modeled avoided
costs may not reflect some of the other complexities
identified above, simply looking at variable fuel and
O&M may be misleading. However, basic approaches
using system averages, time-dependent methods, dis-
placement curves, and load dispatch curve analysis can
give reasonable estimates of the impacts of clean energy.
System Averages
The simplest approach to estimating the impacts of the
displaced unit, absent any detailed information on the
system, is to use the average generating unit as a proxy.
Some studies have used this approach. The average sys-
tem costs and the average emissions characteristics can
be used to estimate impacts; however, most analysts
recognize that some types of generating units are al-
most never on the margin and therefore should not be
included in the characterization of the marginal unit.
For example, nuclear units, hydropower, and renewable
resources are very rarely on the margin and unlikely
to be displaced by clean energy sources in the short
run. Moreover, the average cost of generation can differ
greatly from the marginal source of generation.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 65
-------
In response to this observation, one approach some-
times used is to characterize the remaining units
specifically, the fossil unitsas a representation of the
average marginal unit. This is an improvement over the
system average, but still does not capture the potential
impact of a variety of clean energy resources, each with
differing impact patterns. For example, in many re-
gions of the country coal units are on the margin only
a small number of hours during the year. Thus, using
a fossil average may understate cost savings and over-
state emissions impacts of the clean energy resource.
Despite these limitations, absent any detailed informa-
tion on the impact of the resource or the nature of the
marginal generation, this approach is an option.
Time Dependent Methods
Another method to estimate the impacts of clean ener-
gy resources, including effects on costs and emissions,
is to identify those resources that are expected to be
displaced depending on the time the clean energy im-
pacts occur. The most detailed approach is to identify
the marginal generating unit on an hourly basis. Clean
energy impacts (in kWh) can then be mapped (using
the time of impact estimates described above) to the
appropriate marginal generation source. Costs savings
(and emissions impacts) can then be estimated.
Time-dependent methods do not need to be on an
hourly basis; several less data-intensive basic approach-
es (displacement curves and load curve analysis) are
available and described below:
Displacement Curves
Another approach to estimating what will be displaced
by clean energy involves displacement curves. Baseload
plants operate all of the time throughout the year
because their operating costs are low and because they
are typically not suitable for responding to the many
fluctuations in load that occur throughout the day. As a
result, they would not be expected to be displaced with
any frequency. These plants would have high capacity
factors (e.g., greater than 0.8). Capacity factor is the
ratio of how much electricity a plant produces to how
much it could produce, running at full capacity, over a
given time period. Load-following plants, in contrast to
baseload plants, can quickly change output, have much
lower capacity factors (e.g., less than 0.3) and are more
likely to be displaced.
A displacement curve can be developed to identify
what generation is likely to be displaced. The curve
would reflect the likelihood of a unit being displaced,
FIGURE 3.2.2 DISPLACEMENT CURVE BASED
ON CAPACITY FACTOR
Sample curve for relating displacement to capacity factor
1 nnฐ/ ~ ~
:c
.c
"tU OJ
u m 60% -
Q] d-
og
.*** rts 4flCฃ.
* ฃ w%
^ E
*o i? ^ncjt. -
g ^u%
V
y
nฐฃ, -
\
\
\
\
\.
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Lnit Capacity Factor
Source: Keith and Biewald, 2005.
based on a proxy for its place in the dispatch order. A
reasonable proxy for the likelihood of a generating unit
to be displaced by a clean energy measure is the units
capacity factor. Figure 3.2.2 illustrates this concept
using capacity factor as a proxy. Baseload plants on
the right side of the curve, such as nuclear units, are
assumed to be very unlikely to be displaced; peak load
plants on the left, such as combustion turbines, are
much more likely to be displaced. These capacity factor
estimates can be based on an analysis of actual dispatch
data, modeling results, or judgment. Historic data on,
or estimates of, capacity factors for individual plants are
available from EPA's eGRID database (http://www.epa.
gov/deanenergy/'energy-resources/egrid/index.html).
It is important to note that a displacement curve may
not capture some aspects of electric system operations.
For example, an extended outage at a baseload unit
(for scheduled maintenance or unanticipated repairs)
would increase the use of load-following and peaking
units, affecting the change in net emissions from the
clean energy project. According to the displacement
curve, this plant would be more likely to be displaced,
even though it would rarely if ever be on the margin.
The relationship between capacity factor and percent of
time it will be displaced could be determined analyti-
cally (e.g., examining historical data on the relationship
between a unit's capacity factor and the time it is on the
margin. More likely a judgment could be made about
this relationship. Other proxies could serve to de-
velop this curve, including unit type (e.g., coal steam,
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 66
-------
nuclear, combustion turbine), heat rate, or pollution
control equipment in place.
Load Curve Analysis
In general, generating units are dispatched in a predict-
able order that reflects the demand on the system and
the cost and operational characteristics of each unit.
These plant data can be assembled into a generation
"stack," with lowest marginal cost units on the bot-
tom and highest on the top. A dispatch curve analysis
matches each load level with the corresponding mar-
ginal supply (or type of marginal supply). Table 3.2.5,
Hypothetical Load for One-Week Period, and Figure
3.2.3, a hypothetical dispatch curve representing 168
hours by generation unit, ranked by load level, provide
a combined example of a dispatch curve that represents
168 hours (a one-week period) during which a hypo-
thetical clean energy resource would be operating.
Table 3.2.5 illustrates this process for a one-week pe-
riod. There are 10 generating units in this hypothetical
power system, labeled 1 through 10. Column [3] shows
the number of hours that each unit is on the margin.
In many cases, dispatch curves are available from the
local power authorities and Load Balancing Authorities
[e.g., a regional Independent System Operator (ISO)].
If this information is not available, states can attempt to
construct their own analysis.
Constructing a dispatch curve requires data on:
Historical utilization of all generating units in the
region of interest;
Operating costs and emission rates (to support
emissions estimation, as described in Chapter 4) of
the specific generating units, for the most disaggre-
gate time frame available (e.g., seasonally, monthly);
Energy transfers between the control areas of the
region and outside the region of interest (because
the marginal resource may be coming from outside
the region); and
Hourly regional loads.
Operating cost and historical utilization data can
typically be obtained from the EIA (http://www.eia.
doe.gov/cneaf/electricity/page/data.html) or the local
Load Balancing Authority5 When generator cost data
5 Often these sources can also provide generator-specific emission rates for
estimating potential emission reductions from clean energy.
FIGURE 3.2.3 A HYPOTHETICAL LOAD
DURATION/DISPATCH CURVE REPRESENTING
168 HOURS (shown in half-day increments)
by generation unit, ranked by load level
DOil Combustion Turbine, Old
Gas Combustion Turbine
Oil Combustion Turbine, New
D Gas Steam
Oil Steam
DGas Combined Cycle, Typical
Gas Combined Cycle, New
DCoal, Typical
DCoal, New
Nuclear
1 13 25 37 49 61 73 85 97 109 121 133 145 157
Hour
Source: Developed by Synapse Energy, unpublished, 2007.
TABLE 3.2.5 HYPOTHETICAL LOAD FOR ONE-
WEEK PERIOD: HOURS ON MARGIN AND
EMISSION RATE
1
2
3
4
5
6
7
8
9
10
Oil Combustion Turbine, Old
Gas Combustion Turbine
Oil Combustion Turbine, New
Gas Steam
Oil Steam
Gas Combined Cycle, Typical
Gas Combined Cycle, New
Coal, Typical
Coal, New
Nuclear
5
10
9
21
40
32
17
34
0
0
Weighted average, SO2 emissions (Ibs/MWh): 5.59
CHAPTER 3 | Assessing the Multiple Benefits of Clean Energy 67
-------
are not available, capacity factors (from the eGRID
database, for example, as described above) for tradi-
tional generating units can be used to approximate the
relative cost of the unit (those with the highest capacity
factors are assumed to have the lowest cost). As an
exception, variable power resources such as wind and
hydropower are assumed to have lower costs than fossil
fuel or nuclear units.
Operational data (or simplifying assumptions) regard-
ing energy transfers between the control areas of the
region and hourly regional loads can be obtained from
the ISO or other Load Balancing Authority within the
states region.
Dispatch curve analysis is commonly used in plan-
ning and regulatory studies. It has the advantage of
incorporating elements of how generation is actually
dispatched while retaining the simplicity and transpar-
ency associated with non-modeling methods. However,
this method can become labor-intensive relative to
other non-modeling methods for estimating displaced
emissions if data for constructing the dispatch curve
are not readily available. Another disadvantage is that
it is based on the assumption that only one unit will be
on the margin at any given time; this generally is not
true in most regions.
Methods described earlier, such as displacement
curves, can support the development of a simplified
dispatch curve. For example, capacity factors can be
used to "fill" the horizontal segments on the curve as
shown in Figure 3.2.3. One can assume that units with
capacity factors greater than 80 percent can fill the
baseload segments and that peaking units, with the
lowest capacity factors, would fill the peak segments.
Units with capacity factors between 80 and 60 percent
would fill the next slice of the dispatch curve, and so
on. The resolution would reflect available data or the
ability to develop meaningful assumptions. The hope
is that the level of aggregation is such that the units'
characteristics are generally similar and as such the
marginal unit would be approximated by the group av-
erage. If data allows, it is possible to take into account
differences in units that drive their costs and emissions
(e.g., general unit type and burner type, the presence of
pollution control equipment, unit size, fuel type).
Step 3: Identify the operating costs of marginal units
to be displaced. This process varies depending on
whether the market is regulated or restructured.
In regulated markets, short-run avoided energy costs
typically include fuel costs, a variable O&M cost, and
marginal emissions costs for the highest-cost generator
in a given hour. Data sources for control area hourly
marginal costs include the U.S. Federal Regulatory
Commission (FERC) form 714 (http://www.ferc.gov/
docs-filing/eforms/form- 714/overview, asp).
In restructured markets, where RTOs administer re-
gional wholesale power markets, economic dispatch
is conducted on the basis of bid prices rather than
generators' marginal costs (theoretically equivalent to
the marginal cost). This information is available at each
ISO's Web site (see Information Resources at the end of
this chapter for the Web sites of individual ISOs).
For longer-term analysis it is necessary to forecast
cost increases. Historical hourly operating costs for
the marginal unit (i.e., regulated markets) or market
prices (i.e., restructured markets) can be escalated
using forward market electricity prices, though the
forecast time frame is limited. Forward electricity
prices are available from energy traders and industry
journals such as Platt's Mega Watt Daily (http://www.
platts.com/Electric%20Power/Newsletters%20&%20/
Megawatt%20Daily/).
Step 4: Calculate the short-run avoided costs of elec-
tricity generation. For each hour or time of use period,
multiply the cost of the marginal unit or hourly energy
market price by the reduction in load (for demand-
side resources) or the increase in generation (for
supply-side resources), as estimated using techniques
described in Chapter 2. Typically, avoided costs are
expressed as the annual sum of these avoided costs for
each hour or other time period.
The Estimating Short-Run Avoided Cost text box illus-
trates how all four steps can be used to estimate short-
run avoided costs.
Key Considerations
These basic methods have some limitations that should
be considered when choosing an approach:
Methods that rely on historical data are limited to
replicating what occurred in the past. Substantial
changes in costs or performance of generation, or
other restrictions on their operations (e.g., climate
legislation, requirements for a renewable portfolio
standard) could fundamentally change the opera-
tion of the system and the implied dispatch curve.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 68
-------
Estimating Short-Run
Avoided Cost
To illustrate the described approach
for estimating short-run avoided costs,
consider the case of a state that wishes
to evaluate the potential benefits of
an energy efficiency program. Sample
calculations are illustrated in the ac-
companying table.
Step 1: The state estimates that the en-
ergy efficiency program would reduce
electricity demand as shown in the
Avoided Electricity column (based on
an analysis of annual savings from the
typical system and a typical load shape).
Step 2: Using a load curve analysis,
the state estimates that natural gas
combustion turbines are typically on
the margin during peak periods for both
summer and winter, a mix of natural gas
combined cycle units and natural gas-fired
steam units (about 50% of each) are on the
margin during shoulder periods, and exist-
ing coal-fired generators (pulverized coal)
are typically on the margin during the off-
peak periods.
SAMPLE CALCULATION OF SHORT-RUN ENERGY AVOIDED COSTS
Avoided Cost for Time Total Avoided
Time Period Electricity (MWh) Period ($/kWh) Energy Cost ($)
Summer Peak (912 hours)
Summer Shoulder (1368 hours)
123,120
153,900
0.08
0.06
Summer Off-Peak (1368 hours) 20,520 0.03
Winter Peak (1278 hours)
Winter Shoulder (1917 hours)
Winter Off-Peak (1917 hours)
Total
115,020
143,775
19,170
575,505
0.07
0.06
0.03
9,234,000
8,772,300
513,000
8,051,400
8,195,175
479,250
35,245,125
Step 3: The avoided costs associated with
each of these marginal generating tech-
nologies are estimated based on typical
variable operating and fuel costs for those
types of units estimated to be on the mar-
gin. The results are show in the Avoided
Energy Cost for Time Period column.
Step 4: The Total Avoided Energy Cost
column shows the result of multiplying the
Avoided Electricity column by the Avoided
Energy Cost for Time Period column. Sum-
ming across all periods yields the expected
avoided costs for one year.
Even without such fundamental changes, the
system changes over time as new units are
added, existing units are retired, and units shift
in dispatch order. Analyses based on histori-
cal data do not capture these shifts, so to the
extent that estimates are being developed for
the future these types of basic methods must
be used with caution.
These methods may not adequately address the is-
sue of leakagein which increases in clean energy
result in reductions in generation outside the re-
gion of interest (e.g., in another state or region)if
these transactions are not explicitly accounted for
in the analysis.
Sophisticated Methods for Estimating Short-Run
Avoided Costs: Dispatch Modeling
Sophisticated simulation modeling, such as electric
dispatch modeling, requires developing a detailed rep-
resentation of the electric system with many individual
input assumptions. While developing a full input data
set for a dispatch simulation model can be a resource-
intensive task, the output from a simulation model can
provide more valid estimates than a basic approach,
especially for clean energy resources with more avail-
ability at certain times and for projections of clean
energy impacts in the future. Dispatch models can also
be employed to develop parameters that can be used
to estimate the impacts of a large range of clean energy
resources. For example, multiple model runs can be
performed estimating impacts of changes in genera-
tion requirements at certain seasons and times of day
(e.g., winter peak, summer peak, winter base, etc.).
These parameters, such as the marginal emission rate
and avoided costs, can be applied to estimates of the
impacts of clean energy resources at those same times.
Dispatch models simulate the dynamic operation of
the electric system given the characteristics of specific
generating units and system transmission constraints.
They typically do not predict how the electric system
will evolve but instead can indicate how the existing
electric sector will respond to a particular clean energy
policy or measure. This is appropriate in the short run
when the electric system is more likely to react than to
evolve due to clean energy measures. Dispatch models
specifically replicate least-cost system dispatch and can
be used to determine which generating units are dis-
CHAPTER3 | Assessing the Multiple Benefits of Clean Energy 69
-------
NEW YORK ENERGY $MARTSM PROGRAM COST
EFFECTIVENESS ASSESSMENT
The New York State Energy Research and Development
Authority (NYSERDA) periodically evaluates the cost-
effectiveness (using a benefit-cost ratio) of New York Energy
Smart energy efficiency programs. NYSERDA uses a production
costing model, MAPS, to forecast the avoided energy and
capacity benefits of the programs for several years. Avoided
energy costs are forecasted by applying MAPS escalation rates
to the weighted average energy price by location and time
period. The weighted average energy prices are based on
historical hourly NYISO day-ahead market data for January
2000 through December 2004. The avoided capacity costs
are forecasted by applying the same escalation rates to NYISO
monthly capacity data by location and time period.
Source: Heschong Mahone Group, Inc., 2005.
placed and when they are displaced based on economic
and operating constraints.
Hourly dispatch modeling is generally used for near-
term, highly detailed estimations. This approach is ap-
propriate for financial evaluations of specific projects,
short-term planning, and regulatory proceedings. Sen-
sitivity cases can be run to explore the range of possible
impact values. While this type of modeling is generally
seen as very credible in these contexts, it often lacks
transparency. For example, dispatch models vary in
terms of how they treat outage rates, heat rates, bidding
strategies, transmission constraints, and reserve mar-
gins. Underlying assumptions about these factors may
not be apparent to the user. Moreover, labor and data
needs are extensive. Software license and labor costs
can be prohibitively high for many agencies and stake-
holders, who often must rely on the results of dispatch
modeling conducted by utilities and their consultants
for regulatory proceedings.
Generally, this method involves modeling electricity
dispatch with and without the new resource, on an
hourly basis, for one to three years into the future. As
with basic estimation methods, it is essential to estab-
lish the specific operational profile of the clean energy
resource. Alternatively, an hourly dispatch model can
be used to determine hourly marginal costs and emis-
sion rates (Ibs/kWh), which can then be aggregated
by time period and applied to a range of clean energy
resources according to their production characteristics.
Some models, described later in this chapter, simulate
both capacity planning and dispatch, although they
may have a simpler representation of dispatch (e.g.,
seasonally, with multiple load segments). These models
are applied similarly to models that strictly address
dispatch, but offer the ability to capture the differing
marginal resources over load levels and time.
Tools
There are several dispatch models available for states
to use:
EnerPrise Market Analytics (powered by PROSYM)
supported by Ventyx*.
A chronological electric power production costing
simulation computer software package, PROSYM
is designed for performing planning and op-
erational studies. As a result of its chronological
nature, PROSYM accommodates detailed hour-
by-hour investigation of the operations of electric
utilities. Inputs into the model are fuel costs, vari-
able operation and maintenance costs, and startup
costs. Output is available by regions, by plants,
and by plant types. The model includes a pollution
emission subroutine that estimates emissions with
each scenario, http://wwwl.ventyx.com/analytics/
market-analytics.asp
Multi-Area Production Simulation (MAPS) devel-
oped and supported by GE Energy and supported
by other contractors.
A chronological model that contains detailed
representation of generation and transmission
systems, MAPS can be used to study the impact on
total system emissions that result from the addi-
tion of new generation. MAPS software integrates
highly detailed representations of a system's load,
generation, and transmission into a single simula-
tion. This enables calculation of hourly production
costs in light of the constraints imposed by the
transmission system on the economic dispatch of
generation. http://www.gepower. com/prod_serv/
products/utitity_software/en/ge_maps/index.htm
* Plexosfor Power Systems owned by Energy
Exemplar.
A simulation tool that uses LP/MIP (Linear
Programming/Mixed Integer Programming) opti-
mization technology to analyze the power market,
Plexos contains production cost and emissions
modeling, transmission modeling, pricing model-
ing, and competitiveness modeling. The tool can be
used to evaluate a single plant or the entire power
system. http://www. energyexemplar. com
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 70
-------
PowerBase Suite (including PROMOD IV*) sup-
ported by Ventyx.
A detailed generator and portfolio modeling
system, with nodal locational marginal pricing
forecasting and transmission analysis, PROMOD
IV can incorporate extensive details in generating
unit operating characteristics and constraints,
transmission constraints, generation analysis, unit
commitment/operation conditions, and market
system operations, http://wwwl.ventyx.com/
analy tics/promod. asp
3.2.1.b Avoided Costs of Power Plant
Capacity
While the avoided cost of energy generation is the
major short-run benefit, avoided costs of power plant
capacity in the long run can be significant and should
be included in resource decisions.6 For example, in the
short run, surplus centralized generation capacity that
is freed up by clean energy policies and programs can
be sold to other utilities in the region for meeting their
capacity needs. These costs are based on the levelized7
6 For more information about establishing energy efficiency as a high priority
resource in long run planning, see National Action Plan for Energy Efficiency
Vision for 2025: A Framework for Change, November 2008. http://www.epa.
gov/deanenergy/energy-programs/napee/resources/vision2025.html.
7 The present value of capital costs, levelized in real dollars to remove the
effect of inflation.
capital costs of peaking capacity (e.g., a combustion
turbine) or on the market price for peaking capac-
ity. This is a critical factor in competitive wholesale
markets. Over the long run, however, new clean energy
initiatives typically avoid or defer both the cost of
building new power plants and the cost of operating
them. These are the avoided costs of power plant capac-
ity that can be estimated using either basic estimation
or sophisticated simulation approaches.8 Both have
advantages and limitations, as described in Table 3.2.6.
Basic Methods for Estimating Avoided Costs of
Power Plant Capacity
Basic estimation methods involve the use of tools such
as spreadsheets to estimate any long-run avoided costs
of power plant capacity that may result due to a clean
energy measure under consideration. One method
commonly used is the proxy plant approach. This ap-
proach involves estimating the avoided cost of a power
plant that might be built in the future. Energy cost
estimates (as described above) would reflect this plant's
dispatch costs for future estimates and the capital costs.
Depending on future expectations of capital costs,
fuel prices, and environmental requirements, either a
8 For information about how utilities estimate avoided costs, see The Guide
to Resource Planning with Energy Efficiency: A Resource of the National Ac-
tion Plan for Energy Efficiency, November 2007, www.epa.gov/cleanenergy/
documents/resource planning.pdf, or Costing Energy Resource Options: An
Avoided Cost Handbook for Electric Utilities (Tellus Institute, 1995).
TABLE 3.2.6. COMPARISON OF BASIC AND SOPHISTICATED APPROACHES FOR QUANTIFYING AVOIDED
COSTS OF POWER PLANT CAPACITY
Peaker construction cost.
See also above for combined
capacity & energy estimate.
Simple.
May already be available.
Peaker methodology does
not reflect opportunities to
displace baseload in the long
Rough estimates.
Preliminary screening of
demand response resources
Overview-type policy
assessments.
Sophisticated approach
Capacity Expansion/Ventyx.
PowerBase Suite.
-------
ELECTRIC ENERGY EFFICIENCY AND RENEWABLE ENERGY IN
NEW ENGLAND: THE OTC WORKBOOK
An analysis conducted by the Regulatory Assistance Project
(RAP) explains how energy efficiency and renewable energy
have led to many positive effects on the general economy,
the environment, and energy security in New England while
also quantifying these effects in several new ways. The report
assesses the air quality effects of efficiency and renewable
investments using the OTC Workbook tool. The analysis finds
that there is clear progress in reducing CO2 emissions from the
deployment of energy efficiency and renewable energy. The
projections by the OTC Workbook indicate that due to current
energy efficiency programs, 22.5 million tons of CO2 emissions
are avoided from 2000-2010.
Source: The Regulatory Assistance Project, http://www.raponline.org/
Pubs/RSWS-EEandREinNE.pdf
combined cycle combustion turbine or a new advanced
coal plant may be used as the proxy plant to represent
the long-run avoided costs of energy and capacity of
clean energy initiatives.
Data required for this method include:
Cost and performance information for the proxy
plant; and
Capital cost escalation rates, a discount rate, and
other financial data.
Utilities are one possible source of these data and often
provide this information to public utility commissions
in resource planning and plant acquisition proceed-
ings. Other data sources include:
Regional transmission organizations, independent
system operators, and power pools. These sources
maintain supply and demand projections by region
and often sub-region.
The U.S. Energy Information Administration (EIA)
Annual Energy Outlook. This resource provides
long-term projections of fuel prices and electricity
supply and demand. In addition, some states and
regions develop their own forecasts of electricity
demand, fuel prices, and other variables, http://
www.eia.doe.gov/oiaf/aeo/
Regional reliability organizations. These organiza-
tions can provide information on required reserve
margins.
A RESOURCE FOR CALCULATED AVOIDED EMISSIONS:
THE MODEL ENERGY EFFICIENCY PROGRAM IMPACT
EVALUATION GUIDE
The Model Energy Efficiency Program Impact Evaluation Guide
provides guidance on model approaches for calculating energy,
demand, and emissions savings resulting from energy efficiency
programs. The Guide is provided to assist in the implementation
of the National Action Plan for Energy Efficiency's five key
policy recommendations and its Vision of achieving all cost-
effective energy efficiency by 2025. Chapter 6 of the report
presents several methods for calculating both direct onsite
avoided emissions and reductions from grid-connected electric
generating units. The chapter also discusses considerations for
selecting a calculation approach (NAPEE, 2007).
The Bureau of Economic Analysis (BEA). The BEA
provides information on economic forecasts. The
BEA releases measures of inflation (e.g., the Gross
Domestic Product Implicit Price Deflator), which
are available on its Web site http://www.bea.gov/
national/index.htm#gdp
The Securities and Exchange Commission (SEC)
and the Federal Energy Regulatory Commission
(FERC). Individual utility historical financial data
are available in annual reports and other utility
filings with the SEC and FERC. Utilities file annual
10-K and quarterly 10-Q company reports with
the SEC. These data are available from the SEC
EDGAR system at http://www.sec.gov/edgar.shtml
Utilities also file FERC Form 1, which is available
from FERC at http://www.ferc.gov/docs-fiHng/
eforms/farm-1/viewer-instruct.asp. They can also
be retrieved from the eLibrary at http://www.ferc.
gov/docs-filing/eli brary. asp.
Using data on initial construction costs, fixed and vari-
able operating costs, and financial data, a discounted
cash flow analysis can be conducted. Once estimated,
the net present value of the cost of owning the unit that
reflects the full carrying costs of the new unit (includ-
ing interest during construction, debt servicing, prop-
erty taxes, insurance, depreciation, and return to equity
holders) can be converted to annualized costs (in $/
kW-year). The annual capital costs ($/kW-year) can
be multiplied by the annual capacity savings from the
technology to estimate the avoided capital costs. The
load profile information (reductions in demand at peak
hours), discussed earlier would provide an estimate of
displaced capacity, or simpler estimates can be used.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 72
-------
Sophisticated Methods for Estimating Avoided
Costs of Power Plant Capacity: Capacity
Expansion Models
Sophisticated simulation methods, such as capacity ex-
pansion models (also called system planning models),
can be used to quantify the long-run avoided capacity
costs that result from implementing clean energy
measures. Capacity expansion models predict how the
electric system will evolve over time, including what
capacity will be added through the construction of new
generating units and what units will be retired, in re-
sponse to changes in demand and prices. This method
involves allowing the model to predict what will likely
happen to the resource mix based on costs of new
technology, growth, existing fleet of generating assets,
environmental regulations (current and planned), and
considering dispatch both with and without the new
clean energy resource. Capacity expansion models are
typically used for longer-term studies (e.g., five to 20
years), where the impacts are dominated by long-term
investment and retirement decisions. They are also
typically used to evaluate large geographic areas.
Using capacity expansion models to estimate the
avoided costs of power plant capacity typically involves
the steps described below.
Step 1: Generate a business-as-usual forecast of load
and how it will be met. Some capacity expansion mod-
els use existing generating plants and purchase con-
tracts to serve the load over the forecast period, and the
model (or the modeler) adds new generic plants when
those resources do not meet the load forecast. The type
of plants added depends on their capital and operating
costs, as well as the daily and seasonal time-pattern of
the need for power determined using discounted cash
flow analysis as described earlier. The model repeats
this process until the load is served through the end of
the forecast period and a least-cost solution is found.
This base case contains a detailed schedule of resource
additions that becomes the benchmark capital and
operating costs over the planning period for later use
in the long-run avoided cost calculation.
Step 2: Include the clean energy resource over the plan-
ning period and create an alternate forecast. The fol-
lowing two approaches can be used to incorporate the
clean energy resource into the second projection:
For a more precise estimate of the savings from
a clean energy program, reduce the load forecast
year by year and hour by hour to capture the
Capacity Expansion Modeling involves three steps:
1. Generate a BAU forecast of load, and how load will be met
without the clean energy resources;
2. Create an alternate forecast that includes the clean energy
resources over the planning period to show how load is
expected to be met.
3. Calculate the avoided costs of power plant capacity.
impact of energy efficiency resources, based on the
program design and estimates of its energy and
capacity savings, or add renewable resources as an
available supply. This method would capture the
unique load shape of the clean energy resource.
For a less rigorous estimate (e.g., to use in screen-
ing candidate clean energy policies and programs
during program design), reduce the load forecast
by a fixed amount in each year, proportionally
to load level. This method does not capture the
unique load shape or generation supply of the
clean energy resource.
For renewable resources, add the resource to
the supply mix (or for some models and non-
dispatchable resources, renewable energy could be
netted from load in the same manner as is done for
energy efficiency).
In both the precise and less rigorous methods de-
scribed above, the difference in the projected capital
and operating cost over the planning period of the
two cases is the avoided capacity cost to use in analyz-
ing the clean energy resource. If a per unit avoided
cost, such as the avoided cost per MWh, is needed for
screening clean energy resources or other purposes, it
may be computed by taking the avoided cost (i.e., the
difference between the cost in the two cases) for the
relevant time period (e.g., a given year) and dividing
that by the difference in load between the two cases.
Step 3: Calculate the avoided costs of power plant
capacity. The difference between the costs in the two
projections above represents the annualized or net
present value costs that would be avoided by the clean
energy resource.
Capacity expansion or system planning models can
examine potential long-term impacts on the electric
sector or upon the entire energy systemin contrast
to the dispatch models used to assess the avoided costs
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 73
-------
of energy generation, which focus on only the electric
sector. Capacity expansion models that can examine
the potential impacts of programs upon the entire en-
ergy system are generally used for projecting scenarios
of how the energy system will adapt to changes in sup-
ply and demand or to new policies including emissions
controls. They take into account the complex interac-
tions and feedbacks that occur within the entire energy
system (e.g., fuels and emissions markets), rather than
focusing solely upon the electric sector impacts. This
is important because there are tradeoffs at the system
level in the technological and economic feasibility of
fuels and technologies that may not be captured by a
model that focuses solely on a particular aspect of the
electric system. In addition to capturing the numer-
ous interactions, energy system capacity expansion
models can also model dispatch, although often not in
a chronologic, 8760-hour dispatch.9
Tools: Electric Sector-only Capacity Expansion
Models
Commonly used electric sector-only capacity expan-
sion models for calculating long-run avoided costs of
power plant capacity include:
7PM* developed and supported by ICF International.
This model simultaneously models electric power,
fuel, and environmental markets associated with
electric production. It is a capacity expansion and
system dispatch model. Dispatch is based on sea-
sonal, segmented load duration curves, as denned
by the user. IPM also has the capability to model en-
vironmental market mechanisms such as emission
caps, trading, and banking. System dispatch and
boiler and fuel-specific emission factors determine
projected emissions. IPM can be used to model the
impacts of clean energy resources on the electric
sector in the short and long term, http://www.icfi.
com/Markets/Energy/energy-modeling.asp#2
PowerBase Suite (including Strategist*) supported
by Ventyx.
Strategist is composed of multiple application mod-
ules incorporating all aspects of utility planning and
operations. This includes forecasted load modeling;
marketing and conservation programs; production
cost calculations including the dispatch of energy
9 For more information about using capacity expansion models to estimate air
and GHG emissions from clean energy initiatives, please see Section 4.2.2, Step
2: Quantify Air and GHG Emission Reductions from Clean Energy Measures.
resources; optimization of future decisions; non-
production-related cost recovery (e.g., construction
expenditures, AFUDC, and property taxes); full
pro-forma financial statements; and rate design.
http://wwwl.ventyx.com/analytics/strategist.asp
Tools: Whole Energy-Economy System Planning
Models
Energy system-wide models with electricity sector
capacity expansion capability include:
U.S. DOE National Energy Modeling System
(NEMS) is a system-wide energy model that rep-
resents the behavior of energy markets and their
interactions with the U.S. economy. The model
achieves a supply/demand balance in the end-
use demand regions, defined as the nine Census
divisions, by solving for the prices of each energy
product that will balance the quantities producers
are willing to supply with the quantities consum-
ers wish to consume. The system reflects market
economics, industry structure, and existing energy
policies and regulations that influence market
behavior. The Electric Market Model, a module
within NEMS, forecasts the actions of the electric
power sector over a 25 year time frame and is an
optimization framework. NEMS is used to produce
the Energy Information Administration's Annual
Energy Outlook, which projects the U.S. energy
system through 2030 and is used as a benchmark
against which other energy models are assessed.
http://www. eia. doe.gov/oiaf/aeo/overview/
- MARKet Allocation (MARKAL) Model was cre-
ated by the DOE Brookhaven National Laboratory
in the late 1970s, and is now supported by a large
international users group. MARKAL quantifies the
system-wide effects of changes in resource supply
and use, technology availability, and environmen-
tal policy. The MARKAL model determines the
least-cost pattern of technology investment and
utilization required to meet specified demands and
constraints, and tracks the resulting changes in
criteria pollutant and CO2 emissions. This model is
a generic framework that is tailored to a particular
application through the development of energy
system-specific data. MARKAL databases have
been developed by various groups for national,
regional, and even metropolitan-scale applications.
For example, EPA has developed national and
Census-division level databases (http://www.epa.
gov/appcdwww/apb/globalchange/markal.htm).
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 74
-------
MARKAL requires seconds to an hour to run on
a desktop computer, depending on the size of the
database and the options selected. http://www.
etsap.org/markal/main.html
* Energy 2020 is a simulation model that includes
all fuel, demand, and supply sectors and simulates
energy consumers and suppliers. This model can
be used to capture the economic, energy, and en-
vironmental impacts of national, regional or state
policies. Energy 2020 models the impacts of a clean
energy measure on the entire energy system. User
inputs include new technologies and economic
activities such as tax breaks, rebates, and subsidies.
Energy 2020 uses emission rates for NOX, CO2,
SO2, and PM for nine plant types included in the
model. It is available at the national, regional and
state levels, http://www.energy2020.com/
Key Considerations
While capacity expansion or system planning modeling
is generally seen as very credible in long-run contexts, it:
is more resource-intensive than the estimation
methods and
often lacks transparency due to its complexity and
proprietary nature.
It is important to carefully consider key assumptions,
such as fuel price forecasts and retirements, and the
ability to accurately model the complex factors affecting
the system including environmental and other regula-
tory requirements (e.g., renewable portfolio standards).
These assumptions point to the need for model vali-
dation or calibration against actual data or another
projection model.
Most of the models are supported by their developers
or other consultants who have available data sets. Some
studies calibrate against the NEMS-generated Annual
Energy Outlook produced by DOE's Energy Informa-
tion Administration.
3.2.1.C Avoided Transmission and
Distribution Capacity Costs
Clean energy policies and programssuch as custom-
er-sited renewables and clean DG, including CHP
that are sited on or near a constrained portion of the
T&D system, can potentially:
Avoid or delay costly T&D upgrades, construction,
and associated O&M costs, including cost of capi-
tal, taxes and insurance; and
Reduce the frequency of maintenance, because
frequent peak loads at or near design capacity will
reduce the life of some types of T&D equipment.
Deferral of T&D investments can have significant eco-
nomic value. The value of the deferral is calculated by
looking at the present value difference in costs between
the transmission project as originally scheduled and
the deferred project. Most often, the deferred project
will have a slightly higher cost due to inflation and cost
escalations (e.g., in raw materials), but can have a lower
present value cost when the utility discount rate is con-
sidered (which affects the utility's cost of capital). The
difference in these two factors determines the value of
deferring the project.
The avoided costs of T&D capacity vary considerably
across a state depending on geographic region and
other factors. Figure 3.2.4, California T&D Avoided
Costs by Planning Area in 2003, was developed for
the California Public Utilities Commission in 2003.
It illustrates how avoided costs of T&D capacity vary
in California (in $/kW-year) by planning area, utility,
climate zone, and time of day. Using avoided cost esti-
mates based on these differences, rather than on state-
wide system averages, enables states to better target the
design, funding, and marketing of their clean energy
actions (E3 and RMI, 2004; Baskette et al., 2006).
The benefit of avoided T&D costs is often overlooked
or addressed qualitatively in resource planning, because
estimating the magnitude of these costs is typically
more challenging than estimating the avoided costs of
energy generation and plant capacity. For example, the
avoided T&D investment costs resulting from a clean
energy program are highly location-specific and depend
on many factors, including the current system status,
the program's geographical distribution, and trends in
customer load growth and load patterns. It is also dif-
ficult to estimate the extent to which clean energy mea-
sures would avoid or delay expensive T&D upgrades,
reduce maintenance, and/or postpone system-wide
upgrades, due to the complexity of the system.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 75
-------
FIGURE 3.2.4 CALIFORNIA T&D AVOIDED
COSTS BY PLANNING AREA IN 2003
T&D Avoided Cost by Planning
Division ($/kW-yr)
$53.30 to $80.00
11 $26.70 to $53.29
$0.00 to $26.69
Source: Baskette et al, 2006.
The most appropriate approach for estimating avoided
T&D costs is the system planning approach.10 The
system planning approach uses projections and thus
can consider future developments, whether conducted
via a modeling or non-modeling approach. Generally,
it is difficult to be precise when calculating the avoided
cost of T&D capacity because these costs are very
site-specific and their quantification involves detailed
engineering and load flow analyses.
The system planning approach uses projected costs and
projected load growth for specific T&D projects based
on the results from a system planning studya rigorous
engineering study of the electric system to identify site-
specific system upgrade needs. Other data requirements
include site-specific investment and load data. This ap-
proach assesses the difference between the present value
10 Aprojected embedded analysis approach based on historic data also ex-
ists, but is considered appropriate for cost allocation during ratemaking. For
estimating avoided costs due to energy efficiency measures it is important
to consider future capital investment plans, making the system planning ap-
proach preferable.
of the original T&D investment projects and the present
value of deferred T&D projects.11
Another factor affecting location-specific T&D project
cost estimates is system congestion and reliability.
During periods of high congestion, interconnected
resources that can be dispatched at these specific times
are credited at time-differentiated avoided costs. This
approach is used by the California PUC to estimate
long-term avoided costs to support analyses of the
cost-effectiveness of energy efficiency measures. [See
Section 3.5, Case Studies (E3 and RMI, 2004)]. Reli-
ability considerations are reflected in avoided cost
calculations through consideration of the Loss of Load
Probability (LOLP), which is an indicator of the prob-
ability of failure to serve loads (NARUC, 1992).12
Tools
Specialized proprietary models of the T&D system's
operation may be used to identify the location and
timing of system stresses. Examples of such models
include the following:
PowerWorld Corporation offers an interactive power
systems simulation package designed to simulate high
voltage power systems operation on a variable time
frame. http://www.powerworld. com/
Siemens (PSS*E) offers probabilistic analyses and
dynamics modeling capabilities for transmission plan-
ning and operations. https://www.energy.siemens.
com/cms/00000031/en/ueberuns/organizati/services/
siemenspti/softwareso/Pages/psse_1439533.aspx
3.2.1.d Avoided Energy Loss During
Transmission & Distribution
In addition to avoiding electricity generation, power
plant capacity additions, and T&D capacity additions,
clean energy policies and programs can avoid energy
losses during T&D when these resources are located
near the electricity consumer. Avoided energy losses
during T&D can be estimated by multiplying the esti-
mated energy and capacity savings from clean energy
11 The investment in nominal costs is based on revenue requirements that
include cost of capital, insurance, taxes, depreciation, and O&M expenses
associated with T&D investment. (Feinstein et al, 1997; Orans et al., 2001;
Lovins et al, 2002)
12 LOLP can be used to allocate the marginal capacity costs to time periods
(NARUC, 1992, 118). A LOLP of 0.01 means there is a one percent probability
that the utility might not be able to serve some or all of customer load. Because
LOLP increases as customer usage increases, a LOLP-weighted marginal
capacity cost will be high during high LOLP periods.
CHAPTER 3 | Assessing the Multiple Benefits of Clean Energy 76
-------
VERMONT USES SYSTEM PLANNING APPROACH TO
ESTIMATE AVOIDED TRANSMISSION COSTS
The Vermont Electric Company (VELCO) owns and maintains
the bulk transmission facilities in the state to serve all the
electric distribution utilities. In 2003, VELCO undertook a study
of alternatives to a proposed major upgrade in the northwest
corner of Vermont. The transmission upgrade was reliability-
driven and urgently needed, which resulted in a very high
bar for alternatives. VELCO reached an agreement with the
Vermont Department of Public Service to conduct a thorough
study of distributed generation, energy efficiency, and new
central generation as alternatives to the upgrade.
The study identified a range of central generation and
distributed generation options and estimated their costs. In
addition, a location-specific study of the available energy
efficiency potential and the program costs for delivering
that potential was prepared. Various combinations of energy
efficiency and generation were assembled as alternatives to the
proposed transmission project and compared based on total
present value of cost of service. The study determined the cost
of the transmission upgrade and the cost of a smaller upgrade
so that the difference in those two costs could be used to
assess the cost-effectiveness of the alternative resource
package. While the alternatives were not adopted, due in part
to the fact that only the transmission option's costs could be
spread across the whole ISO region, this study demonstrates
one way to use the system planning approach to estimate
avoided transmission costs.
Source: LaCapra Associates, 2003; Orans, 1989; Orans, 1992.
policies and programs located near or at a customer
site by the T&D energy loss percentage. An approach
for determining the energy loss is described below.
The energy loss factor is the percent difference between
the total energy supplied to the T&D system and the
total energy taken off the system for delivery to end-use
customers during a specified time period, calculated as
1 minus (delivered electricity/supplied electricity). T&D
losses in the range of 6 percent to 10 percent are typical,
which means that for every 1 kWh saved at the custom-
er's meter, 1.06-1.10 kWh is avoided at the generator.
Line loss is typically higher when load is higher, es-
pecially at peak times when it can be as great as twice
the average value. The line loss reductions from energy
efficiency, load control, and DG are thus significantly
higher when the benefits are delivered on peak than
when they occur at average load levels, which greatly
enhances the reliability benefits. A clean energy mea-
sure that saves 1.0 KWh of power at the customer's me-
ter may save, for example, 1.2 KWh from the generator
during peak hours simply because line losses are higher
at peak times.
The significance of losses in high load periods is fur-
ther increased by the high marginal energy costs and
energy prices experienced at those times. Due to the
variation in loads over the course of the year, T&D loss
estimates are more precise when developed for short
time periods (e.g., less than one year).
Utilities routinely collect average annual energy loss
data by voltage level (as a percentage of total sales at
that level). RTOs and ISOs also provide loss data. Note
that transmission loss, which is smaller than distribu-
tion loss, may be included in wholesale energy prices
in restructured markets.
Estimates of line loss can be applied to the energy
impacts estimated as described in Chapter 2. If load
profile information is available, then estimates can
reflect the higher on-peak loss rate.
3.2.2 HOW TO ESTIMATE THE SECONDARY
ELECTRIC SYSTEM BENEFITS OF CLEAN
ENERGY RESOURCES
Clean energy policies and programs result in many ad-
ditional electric system benefits that affect the efficiency
of electric systems and energy markets. These secondary
benefits have associated cost reductions, but the meth-
odologies for assessing them are sometimes diverse,
qualitative, and subject to rigorous debate. As described
in Section 3.1, some of the key secondary benefits of
clean energy to electric systems and markets include:
Avoided ancillary service costs;
Reductions in wholesale market prices;
Increased reliability and improved power quality;
Avoided risks associated with long lead-time
investments, such as the risk of overbuilding the
electric system;
Reduced risks from deferring investments in tra-
ditional centralized resources until environmental
and climate change policies take shape; and
Improved fuel diversity and energy security.
The ability to estimate the secondary benefits of clean
energy policies and programs and the availability
of methods vary depending on the benefit. These
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 77
-------
ANCILLARY SERVICES THAT CLEAN ENERGY RESOURCES CAN
PROVIDE TO THE SYSTEM
Operating reserve - Spinning: Generation synchronized to the
grid (i.e., "spinning") and usually available within 10 minutes
to respond to a contingency event. For example, 50 MW of
spinning operating reserve means that a generation unit can
increase its output by 50 MW within 10 minutes.
Operating reserve - Supplemental: Generation that is
available within 30 minutes but is not necessarily synchronized
to the grid.
Reactive Power/Voltage Support: The ability of a generator
to "absorb" or "generate" reactive power to meet voltage
standards on the grid.
methods are less mature than those for primary ben-
efits, and as such tend to rely more upon non-modeling
estimation approaches than more sophisticated simula-
tion modeling ones. Secondary electric system benefits,
and methods for estimating them, are described below.
3.2.2.a Avoided Ancillary Services Costs
"Ancillary services" is a catch-all term for electric
generator functions needed to ensure reliability, as op-
posed to providing power, and include services such as
operating reserves and voltage support.
Operating Reserves
Energy efficiency programs avoid the need for cor-
responding operating reserves (those generation
resources available to meet loads quickly in the event a
generator goes down or some other supply disruption
occurs) and thus avoid the respective costs.
RTOs routinely report market prices for ancillary ser-
vices. In those regions with ancillary service markets,
such as PJM, NYISO, ISO-NE, ERGOT and the Cali-
fornia RTO, services are provided at rates determined
by the markets and thus are easily valued.13 The market
value of a given MW of clean energy short-term re-
serve is equal to the operating reserve price, as posted
by the RTO or ISO on its Web site.
Voltage Support
Voltage support is important to ensure the reliable and
safe operation of electricity-consuming equipment
and the grid. There are few market metrics available
13 There can be opportunity costs associated with provision of operating
reserve. Some regions allow demand response and other clean energy resources
to bid directly into the energy market.
DEMAND RESPONSE COULD IMPROVE PLANT UTILIZATION
AND REDUCE EMISSIONS IN NEW ENGLAND
Compared with other regional control areas. New England has
a small amount of quick-start capacity relative to the regional
peak load. As such, a number of large oil- and gas-fired steam
units that do not have the ability to start quickly must run
constantly to provide reserve capacity. A study conducted for
the New England Demand Response Initiative (NEDRI) used a
production costing model (PROSYM/MULTISYM) to evaluate
how hypothetical aggressive demand response programs
implemented during the summer of 2006 would affect power
plant utilization and net emissions when such programs are
used for reserve capacity. The study found that the demand
response programs could result in more efficient plant
utilization, reducing operation of the steam units, and increasing
operation of efficient combined-cycle units in the region. If no
diesel generators participate in the demand response programs,
the study identified the additional potential for reductions in
NOx, SO2, and CO2 emissions during the summer.
Source: Synapse Energy Economics, 2003.
to estimate the price of voltage support benefits. The
reactive power provisions in Schedule 2 of the FERC
pro forma open access transmission tariff, or an RTO's
equivalent schedule for reactive support, can be used as
a proxy for the avoided cost of voltage support. How-
ever, the Schedule 2 payments are often uniform across
a large region. As a result, they may not capture differ-
ences in the value of these services in load pockets. Al-
ternately, the difference in reliability with and without
the clean energy resource can also give some indication
of voltage support benefits. (See the reliability metrics
discussion in Section 3.2.2.C Increased Reliability and
Power Quality.)
Some clean energy measures can have direct beneficial
effects on avoiding certain voltage support or reactive
power requirements. Reactive power ancillary services
are local in nature, and clean energy policies and
programs that reduce load in a load pocket area can
minimize the need for local reactive power require-
ments. On the other hand, solar and wind resources
may require backup voltage support due to their inter-
mittent nature.
It is important to note that the avoided costs of reactive
power and other ancillary services are typically smaller
than other costs, such as avoided energy, capacity, and
T&D investment. For example, 2003 reactive power
payments were only 0.52 percent of the total costs of
serving load in PJM (Burkhart, 2005).
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 78
-------
3.2.2.b Reduction in Wholesale Market
Clearing Prices
In addition to the benefits of reduced wholesale elec-
tricity costs (i.e., avoided energy and capacity costs
described in Section 3.3), clean energy resources can
reduce the wholesale market clearing price for electric-
ity as a result of decreased demand for electricity, gas,
or both. This can directly benefit both utilities and
consumers.
The methods for estimating short-run wholesale
market price effects involve relatively well-understood
data and are reasonably straightforward to apply. In
contrast, wholesale market price effects over the long
term involve relatively poorly understood relation-
ships, and estimating these price effects can become
quite complex. For this reason, this section presents
the steps involved in estimating the magnitude of the
price effects of resource additions in the near term us-
ing a basic approach. For longer-term forecasts, a more
sophisticated approach such as a dispatch model may
be preferred.
The potential market price decrease attributable to
a particular clean energy resource can be estimated
based on a load curve analysis as follows.
Step 1. Determine the time period for which the calcu-
lation is to be made.
Step 2. Determine the size of the clean energy resource
(and the hourly shape if relevant), typically in MW.
(For more information, see Step 1: Estimate Clean En-
ergy Operating Characteristics in Section 3.2.1.a)
Step 3. Develop a dispatch curve that can be based
upon either generating unit data (i.e., capacity ratings
and operating costs) or market clearing price data
(typically available from the ISO or control area opera-
tor). (For more information, see Step 2: Identify the
Marginal Units to be Displaced in Section 3.2.1.a)
Step 4. Calibrate or validate the calculation for the case
without the clean energy resource.
Step 5. Analyze a case with the clean energy resource
by reducing demand or adding supply to represent the
clean energy resource.
Step 6. Compare the wholesale market price results for
the two cases. The difference is the wholesale market
PRICE EFFECTS OF DEMAND RESPONSE IN THE NORTHEAST
IN JULY AND AUGUST, 2006
In all four of the structured, RTO-run eastern spot electricity
markets, historically high peak load values occurred during a
week-long heat wave in August 2006. Market coordinators
from New York (over 1,000 MW of load reduction), PJM (520
MW of peak reduction) and New England (625 MW of peak
reduction) all acknowledged the role that demand response
played in keeping peak load lower than what otherwise would
have occurred.
For example, PJM estimated that wholesale prices would have
been $300/MWh higher without demand response during the
highest demand hours of the heat wave, corresponding to a
reported savings of about $650 million for energy purchasers.
Payments to all demand response providers totaled only
$5 million; even considering the potential costs of demand
response programs, such as program administration costs, the
benefit-cost ratio is favorable.
Source: PJM, 2006a, PJM, 2006b.
PRICE EFFECTS DUE TO THE NEW YORK ENERGY $MART
PROGRAM
An evaluation of the cost-effectiveness of a portfolio of
programs under NYSERDA's New York Energy Smart public
benefits program estimated the reduction in average wholesale
electricity prices over the period 2006 (full implementation of
program) to 2008 (the year after which no currently known
planned new capacity is assumed to come online). The
analysis used a production cost model, Multi Area Production
Simulation Software (MAPS), to compare the average annual
wholesale electricity commodity prices in two cases: one with
the New York Energy $martSM Program (the base case), and a
one without the program benefits (the sensitivity case). The
study estimated electricity market price reductions of about
$11.7 million in 2003 to $39.1 million (in 2004 dollars) in 2023
as a result of the program.
Source: Heschong Mahone Croup, 2005.
price reduction benefit (expressed in $/MWh or total
dollars for the time period).
This approach for calculating the market price change
can be applied to the electric energy market and capac-
ity market, if one exists in the region. This benefit can
be calculated using spreadsheets, an electric system
dispatch model (e.g., MAPS, ProSym), or an energy
system model for a more aggregated estimate. Another
approach, used by the CPUC in California's avoided
cost proceeding, is to use historical loads and prices
(CPUC, 2006).
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 79
-------
RELIABILITY CONCEPTS
Reliability refers to the electric system's availability to
consistently serve the demanded load.
Power Quality refers to the consistency of voltage of electricity
supplied to electrical equipment (usually meaning the voltage
stays within plus or minus 5 percent).
RELIABILITY INDICES
SAIFI (system average interruption frequency index): the
average frequency of sustained interruptions per customer
over a predefined area. It is calculated as the total number
of customer interruptions divided by the total number of
customers served.
SAIDI (system average interruption duration index): commonly
referred to as customer minutes of interruption or customer
hours, it provides information on the average time customers
are interrupted. It is calculated as the sum of the restoration
time for each interruption event times the number of
interrupted customers for each interruption event divided by
the total number of customers.
CAIDI (customer average interruption duration index):
the average time needed to restore service to the average
customer per sustained interruption. It is calculated as the sum
of customer interruption durations divided by the total number
of customer interruptions.
MAIFI (momentary average interruption frequency index):
considers momentary interruptions resulting from each
single operation of an interrupting device, such as a recloser.
It is calculated as the total number of customer momentary
interruptions divided by the total number of customers served.
RELIABILITY BENEFITS OF CLEAN ENERGY
Clean energy provides reliability benefits because when a
small clean energy unit fails, the result is less catastrophic
than when one large, traditional generating unit fails. For
example, suppose a utility has the choice of installing one
hundred kilowatts of clean DG around its system or installing a
single 10 megawatt generator (100 units times 100 kW). In this
situation, there would likely be a greater probability of the 10
MW generator being out of service than of finding all 100 of the
smaller units out of service. Such an effect can either reduce
the reserve margin required (which benefits both the utility and
consumers) or, if the reserve margin is fixed, reduce the price of
reserve capacity (Lovins et al., 2002).
THE IMPORTANCE OF POWER QUALITY
It is important to maintain consistent power quality; otherwise,
electrical equipment can be damaged. For example, consumer
and commercial electrical and electronic equipment is usually
designed to tolerate extended operation at any line voltage
within 5 percent nominal, but extended operation at voltages
far outside that band can damage equipment or cause it to
operate less efficiently.
3.2.2.C Increased Reliability and Power
Quality
An expansion in the use of clean energy resources can
improve both the reliability of the electricity system
and power quality. For example, California's invest-
ments in energy efficiency, conservation, and demand
response played a role in averting rolling blackouts in
the summer of 2001. Power quality problems occur
when there are deviations in voltage level supplied
to electrical equipment. Some forms of clean energy
resources, such as fuel cells, can provide near perfect
power quality to their hosts.
Reliability Metrics
Although clean energy resources can improve system
reliability, measuring these benefits can be difficult. The
most common reliability metrics are indices, which
are relatively well-established and straightforward to
calculate (see text box, Reliability Indices). Historical
reliability data are often available.
Converting reliability benefits into dollar values is
complex, however, and the results of studies that have
attempted to do so are controversial. For this reason,
their use in support of resource decisions is less com-
mon than for other, well-established benefits, such as
the avoided costs of generation, capacity, and T&D.
Power Quality Metrics
The data needed to assess power quality benefits are
neither consistently measured nor comprehensively
collected and reported. Specialized monitoring equip-
ment is typically necessary to measure power defects,
and acceptable standards for power quality have been
changing rapidly.
Power quality improvements produce real economic
benefits for electricity consumers by avoiding damage
to equipment and associated loss of business income
and product, and, in some cases, the need for redun-
dant power supply. At the extreme, some commercial
and industrial processes, such as silicon chip fabrica-
tion and online credit card processing, are so sensitive
to outages or power quality deviations that customers
take proactive steps to avoid these concerns, including
construction of redundant transmission lines or install-
ing diesel or battery backup power. The costs of such
equipment could also be used to estimate the value of
increased reliability and power quality.
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 80
-------
3.2.2.d Avoided Risks Associated with Long
Lead-time Investments Such as the Risk of
Overbuilding the Electric System
Clean energy options provide increased flexibility to
deal with uncertainty and risk related to large, tradi-
tional fossil fuel resources, including:
Clean energy resources, such as wind and photo-
voltaics, reduce the impact on electric system costs
from fuel price uncertainty relative to traditional
resources, and lower the financial risks and costs
associated with generation.
In terms of resource planning, clean energy op-
tions offer great flexibility. If one is unsure that
long-term forecasts for load growth are 100
percent accurate, then clean energy resources offer
greater flexibility due to their modular nature and
relatively quick installation times relative to tradi-
tional resources.14
Clean energy resource options provide more time
to develop technologically advanced, less polluting,
more efficient, large-scale technologies.
All other things being equal, a resource or resource
plan that offers more flexibility to respond to changing
future conditions is more valuable than a less flexible
resource or plan. Techniques such as decision tree
analysis or real option analysis provide a framework
for assessing this flexibility. These approaches involve
distinguishing between events within one's control (i.e.,
decision nodes) and those outside of ones control (i.e.,
exogenous events) and developing a conceptual model
for these events as they would occur over time. Specific
probabilities are generally assigned to the exogenous
events. The results of this type of analysis can include
the identification of the best plan on an expected value
basis (i.e., incorporating the uncertainties and risks) or
the identification of lower risk plans.
Above and beyond the expected value of the plan,
certain resources may have some "option value" if they
allow (or don't foreclose) other resource options in the
future. For example, a plan that involves implementing
some DSM in the near term can have value above its
simple short-run avoided cost, in that it develops the
capability for expanded DSM deployment in the future
if conditions call for it.
14 Of course, clean energy resources carry their own risk of non-performance.
THE IMPORTANCE OF LOW PERFORMANCE CORRELATIONS
Similar resources (e.g., fossil fuels such as coal and oil) tend to
face similar specific risks, and as a result their performances
tend to be correlated. For example, coal and oil both emit
CO2 when burned and thus could be associated with future
climate change regulatory risk, which in turn would likely
increase costs and affect the performance of oil- or coal-fired
generation. On the other hand, disparate resources (e.g., coal
and wind) have lower performance correlationsand hence
more value for offsetting resource-specific risks within the
portfoliothan resources that have little disparity.
3.2.2.e Reduced Risks from Deferring
Investment in Traditional, Centralized
Resources Pending Uncertainty in Future
Environmental Regulations
Clean energy resources offer planners options for
mitigating current and future environmental regulation
risks. Clean energy can reduce the cost of compliance
with air pollution control requirements. Utilities and
states also see clean energy as a way to reduce their
financial risk from future carbon regulations.
For example, a 2008 study looked at 10 utilities in
the western U.S. and examined how their respective
resource plans accounted for future carbon regulations.
The study found that the majority of the 10 utilities
included aggressive levels of energy efficiency and
renewable energy to reduce carbon emissions. The
study also found that in making these decisions the
utilities did not consider the indirect impacts of future
carbon regulations, such as increased wholesale electric
market price, retirements of conventional generation
plants, and the impact on transmission and distribu-
tion expansion (Barbose et al., 2008).
When comparing new generation options in the face
of potential environmental regulations, some states
and utilities are reducing financial risk by placing a
higher cost premium on traditional resources relative
to clean energy. For example, California has adopted
an $8/ton carbon dioxide greenhouse gas adder to be
used in comparing resources (Johnston et al., 2005; CA
PUC, 2004).
3.2.2.f Improved Fuel Diversity and Energy
Security
Portfolios that rely heavily on a few energy resources
are highly affected by the unique risks associated with
any single fuel source. In contrast, the costs of clean
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 81
-------
FIGURE 3.2.5. NATURAL GAS AND ELECTRICITY PRICES IN NEW ENGLAND
A large portion of New England's electricity is generated from natural gas. Due to this high dependence on one fuel source, and because fuel
represents a large portion of the cost to produce electricity, natural gas and electricity prices are highly correlated.
140
120
100
2003
2004
2005
2006
2007
2008
2009
Sources: El A; ISO NE, summary of monthly data, 2006.
energy resources are not affected by fossil fuel prices
and thus can hedge against fossil-fuel price spikes by
reducing exposure to this volatility.
Diversity in technology can also reduce the likelihood
of supply interruptions and reliability problems. For
example, while geothermal plants can be expensive
to construct, they offer an almost constant supply of
energy and are best suited for baseload generation. Gas
turbines, on the other hand, are relatively inexpensive
to construct and can start quickly, but have a high
operating cost and so are best suited for peaking gen-
eration. Figure 3.2.5 illustrates the relationship between
electricity and natural gas prices in New England.
Two approaches for estimating the benefits of fuel and
technology diversification include market share indices
and portfolio variance.
Market share indices. Market share indices, such as
the Herfindahl-Hirschmann Index and Shannon-
Weiner index, identify the level of diversity as a
function of the market share of each resource.15
These indices are computationally simple and the
15 For more information about these indices, see U.S. Department of Justice
and the Federal Trade Commission, Issued April 1992; Shannon, C.E. "A
mathematical theory of communication." Bell System Technical Journal 27:
379-423 and 623-656, July and October 1948.
data required for the indices (annual state electric-
ity generation by fuel type and producer type) are
readily available from the EIA Form 906 database.16
Use of these indices is appropriate for preliminary
resource diversity assessment and as a state or
regional benchmark. Annual state electricity
generation data by producer type and fuel type are
available.
A limitation of these indices is that decisions on how
to classify resources (e.g., calculating the share of
all coal rather than bituminous and subbituminous
coals separately) can have a large effect on the results.
Another shortcoming is that the indices do not differ-
entiate between resources that are correlated with each
other (e.g., coal and natural gas) and thus can under-
estimate the portfolio risk when correlated resources
are included.
Portfolio Variance. The concept of portfolio theory
suggests that portfolios should be assembled and
evaluated based on the characteristics of the port-
folio, rather than on a collection of individually
assessed resources. Portfolio theory and portfolio
variance measures account for risk and uncertainty
by incorporating correlations between resources
16 EIA Form 906 has been superseded by EIA Form 923. Both data sets are
available at http://www.eia.doe.gov/cneaf/electricity/page/eia906 920.html
CHAPTER 3 | Assessing the Multiple Benefits of Clean Energy 82
-------
when projecting overall portfolio performance,
as measured by the standard deviation of cost or
some other measure of performance. The standard
deviation can be calculated for a number of port-
folios, each with a variety of different resources, to
find portfolios that simultaneously minimize cost
and risk. It is important to acknowledge this inher-
ent trade-off between cost and risk; there is not a
single portfolio that lowers both.
Like market share metrics, portfolio analysis does
not readily incorporate the non-price and qualitative
benefits of fuel diversity, such as energy independence,
which can be a benefit of clean energy. It is safer to
have many smaller, generating resource units that are
located in a variety of locations and do not require fuel
stored on-site than to have one easily targeted large
unit. Also, using domestic clean energy resources to
reduce dependence on foreign fuel sources, such as
imported petroleum, may yield political and economic
benefits by protecting consumers from supply shortag-
es and price shocks. Care should be taken to consider
price as well as factors that are not easily quantified
when choosing among portfolios with different cost-
risk profiles.
3.3 CASE STUDIES
The following two case studies illustrate how assess-
ing the electric system benefits associated with clean
energy can be used in the state energy planning and
policy decision-making process.
3.3.1 CALIFORNIA UTILITIES' ENERGY
EFFICIENCY PROGRAMS
Benefits Assessed
Avoided electricity generation costs
Avoided T&D costs
Avoided environmental externality costs
Avoided ancillary services costs
Reduced wholesale market clearing prices
Clean Energy Program Description
In 2005, the California Public Utilities Commission
(CPUC) approved a new method for calculating
avoided costs for use in evaluating 2006-2008 utility
energy efficiency programs in California.
TABLE 3.3.1 COMPARISON OF OLD AND NEW AVOIDED COST METHODOLOGIES
New Methodology Old Methodology
Avoided Cost
Avoided electricity generation costs
Avoided Electric Transmission &
Distribution Costs
Avoided Natural Gas Procurement
Avoided Natural Gas Transportation &
Delivery
Environmental externality Adders for
Electric and Gas
Reliability adder (Avoided ancillary services
costs)
Price elasticity of demand adder (Reduced
wholesale market clearing prices)
^^^M ^^TC ^^^M
Hourly
Hourly
Monthly
Monthly
Annual value, applied by
hour per implied heat rate
Annual value
Time of use period (on-
vs. off-peak) by month
Utility-specific
Utility, planning area and
climate zone specific
Utility-specific
Utility-specific
System-wide (uniform
across state)
System-wide (uniform
across state)
System-wide (uniform
across state)
Annual
Average
Values
None
None
Statewide
None
None
Source: E3 and RMI, 2004
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 83
-------
FIGURE 3.3.1 COMPARISON OF AVOIDED
COSTS FOR THREE EXAMPLE MEASURES
Weighted Average
Avoided Cost
(Levelized SIMWh)
ป ซ ซ ซ 2 3 2
o~! ro .t. . CD o ro -t*
oooooooo r\
:omparison of A
Air
Conditioning
00 -
00 -
00 -
00 -
n
^
: Generation I
voided Costs f
Measures
Outdoor
Lighting
<
r&D
~
^
:>r 3 Example
Refrigeration
___
-------
$5.4 billion, twice as large as the cost of the programs18
(CPUC, 2005).
For More Information
Energy Efficiency Portfolio Plans and Program
Funding Levels for 2006-2008 - Phase 1 Issues. Cali-
fornia Public Utilities Commission. Interim Opin-
ion. September 22, 2005. http://www.cpuc.ca.gov/
PUBLISHED/FINAL_DECISION/49859.htm
3.3.2 ENERGY EFFICIENCY AND
DISTRIBUTED GENERATION IN
MASSACHUSETTS
Benefit(s) Assessed
Reduction in wholesale market clearing prices
Avoided greenhouse gas (CO2) emissions
Clean Energy Program Description
This study explores the potential price and emissions
benefits of different options to increase distributed gen-
eration and energy efficiency in Massachusetts. The op-
tions include the addition of the following new demand
resources over the baseline scenario through 2020:
photovoltaics (PV),
energy efficiency (EE),
combined heat and power (CHP), and
combined EE and CHP.
Method(s)
The analysis required the development of a reference
case to determine what the wholesale electric prices
and carbon dioxide emissions would be without the
additional clean energy resources. It assumed no rate-
payer-funded investments in demand side management
(DSM) programs beginning in 2007 and so it assumed
energy savings achieved through the end of 2006 remain
constant in the future. The reference case also assumed
no new policies to encourage distributed generation.
18 As a result of the energy efficiency programs, California's investor-owned
utilities project savings of about 7,370 GWh of electricity, 1,500 MW of peak
demand, and 122,000 megatherms of natural gas from 2006 to 2008. Relative
to a base case without the programs, the utilities expect to reduce carbon
dioxide emissions by about 6,600,000 tons the equivalent of the emissions
of about 1.2 million cars over the same period.
FIGURE 3.3.2 REDUCTION IN AVERAGE
ANNUAL WHOLESALE ELECTRIC ENERGY
PRICE FOR MASSACHUSETTS PURCHASES IN
2020 UNDER PV, EE, AND CHP+EE CASES
Incremental 3.5%
price impact of
CHP above EE
EE (3,568 GWh)
CHP+EE (8,026 GWh)
Source: Impacts of Distributed Generation on Wholesale Electric
Prices and Air Emissions in Massachusetts, Synapse Energy
Economics, March 31, 2008.
FIGURE 3.3.3 REDUCTIONS IN REGIONAL
CO2 EMISSIONS IN 2020 UNDER PV, EE, AND
CHP+EE CASES RELATIVE TO REFERENCE
CASE MASSACHUSETTS CO2 EMISSIONS
PV (356 GWh)
EE (3,568 GWh) CHP+EE (8.026 GWh)
Source: Impacts of Distributed Generation on Wholesale Electric
Prices and Air Emissions in Massachusetts, Synapse Energy
Economics, March 31, 2008.
The analysis used the PROSYM simulation model to
determine the potential price and emissions impacts
of the scenarios. The model was used to simulate the
average hourly wholesale market clearing prices and
the regional greenhouse gas emissions (apportioned to
Massachusetts based on GWh load) in 2020 under a
reference case and each of the following four scenarios:
CHAPTER 3 | Assessing the Multiple Benefits of Clean Energy 85
-------
250 MW of incremental PV;
Investment in EE sufficient enough to reduce an-
nual growth of Massachusetts' energy consumption
to 0.6 percent;
750 MW of incremental DG from CHP; and
A combined CHP and EE case.
The scenarios are compared against the reference case
to determine the impacts.
Results
The study projected that the combined effect of the
PV, EE, and CHP would be to virtually eliminate load
growth in Massachusetts.
In terms of impact on wholesale market prices:
the 250MW of PV is expected to displace 356
GW of purchases from the wholesale market and
reduce wholesale market prices by $.033/MWh or
0.4 percent,
EE is expected to reduce prices by 1.6 percent, and
the combined EE and CHP scenario would pro-
duce a 5.1 percent reduction in prices.
These market price changes will affect the wholesale
energy costs paid by Massachusetts customers. Even
though it is expected to achieve the lowest reduction in
market clearing prices, PV is expected to achieve the
largest wholesale market cost savings to Massachusetts
consumers: $65 for every MWh generated by PV. EE is
estimated to reduce costs by $24 for every MWh saved.
The study estimates a savings of $35 per MWh of CHP
generation. The values are different due to the different
load shape profiles for each resource and the timing
(and costs) for when each is likely to be used.
For greenhouse gas emissions, each of the alternative
scenarios would achieve reductions of CO2 emissions
relative to the reference case. The combined EE and
CHP scenario is likely to produce the greatest impact,
with a reduction of 2.4 million short tons CO2 /year in
2020. The majority of these reductions come from EE.
For More Information
Impacts of Distributed Generation on Wholesale
Electric Prices and Air Emissions in Massachusetts,
Synapse Energy Economics, March 31, 2008.
http://www. masstech. org/dg/2008-03-Synapse-
DG-Itnpacts-on-NE.pdf
Information Resources
Resource URL Address
Summary of Rigorous Modeling Tools
EnerPrise Market Analytics (powered by PROSYM)
Multi-Area Production Simulation (MAPS)
Plexos for Power Systems
PowerBase Suite (including Promod IV)
Capacity Expansion available from Ventyx
PowerBase Suite (including Strategist)
IPM available from ICF International
PROSYM
http://wwwl.ventyx.com/analytics/market-
analytics.asp
http://www.gepower.eom/procLsen//products/
utility_software/en/ge_maps/index.htm
http://www.energyexemplar.com
http://wwwl.ventyx.com/analytics/promod.asp
http://wwwl.ventyx.com/products-services.asp
http://wwwl.ventyx.com/products-services.asp
h ttp://www. icfi. com/Marke ts/En ergy/en ergy-
modeling.asp#2
http://wwwl.ventyx.com/analytics/market-
analytics.asp
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 86
-------
Primary Electric System Benefits
Bureau of Economic Analysis
California Database for Energy Efficient Resources (DEER). California Energy
Commission database.
California ISO
California Public Utilities Commission (CPUC) 2006. Interim Opinion: 2006 Update of
Avoided Costs and Related Issues Pertaining to Energy Efficiency Resources. Decision
06-06-063 June 29, 2006
E3 and RMI, Methodology and Forecast of Long Term Avoided Costs for the Evaluation
of California Energy Efficiency Programs, October 26, 2004
EIA Annual Energy Outlook
EIA Form EIA-860 (Annual generator data)
EIA Form EIA-861
EIA Form EIA-906 and 920 (power plant database) - now EIA-923
FERC Form 1
FERC Form 714 (control area info)
FERC Form 423 (cost and quality of fuels)
Handy-Whitman 2006. Handy-Whitman Index of Public Utility Construction Costs, a
plant cost index that has been published semi-annually since the 1920s, is published
by Whitman, Requardt & Associates, LLP.
Independent System Operators/ Regional Transmission Organizations
ISO New England
Keith, G., B. Biewald and D. White 2004. Evaluating Simplified Methods of Estimating
Displaced Emissions in Electric Power Systems: What Works and What Doesn't.
Midwest ISO
NYISO
NYMEX
Platt's MegaWatt Daily publishes forward electricity market prices through this paid
subscription newsletter.
http://www.bea.gov
http://www.energy.ca.gov/deer
h ttp://oasis. caiso.com/
http://www.cpuc.ca.gov/PUBLISHED/
COMMENT_DECISION/56572.htm#P86_2251
http://www.ethree.com/CPUC/E3_Avoided_
Costs_Final.pdf
http://www.eia.doe.gov/oiaf/aeo/index.html
http://www.eia.doe.gov/cneaf/electricity/page/
eia860.html
http://www.eia.doe.gov/cneaf/electricity/page/
eia861.html
http://www.eia.doe.gov/cneaf/electricity/page/
eia906_920.html
http://www.ferc.gov/docs-filing/eforms/form-
1/viewer-instruct.asp
http://www.ferc.gov/docs-filing/eforms/form-
714/overview.asp
http://www.eia.doe.gov/cneaf/electricity/page/
eia423.html
http://www.business-magazines.com/
prd!35331.php?siteid = global_BMS_product
http://www.iso-ne.com/
http://www.synapse-energy.com/Downloads/
SynapseReport.2004-ll.CEC-.Evaluating-
Simplified-Methods-of-Estimating-Displaced-
Emissions.04-62.pdf
http://www.midwestiso.org/home
http://www.nyiso.com/public/index.jsp
http://www.nymex.com/index.aspx
http://wwwl.platts.com/Electric%20Power/
Newsletters%20&%20Reports/Megawatt%20
Daily/
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 87
-------
PJM
Portfolio Management: Tools and Practices for Regulators, prepared for the national
Association of Regulatory Utility Commissioners (NARUC), July 17, 2006.
SEC 10K filings.
State regulatory commission rate base and fuel clause adjustment filings
The Massachusetts DG Collaborative Benefits and Costs of Distributed Generation
website compiles a comprehensive list of studies regarding costs and benefits of
distributed generation and distribution planning including the analysis conducted by
the Massachusetts DG Collaborative and Navigant Consulting Inc.
This Excel lookup table contains distribution system deferral values for each of the
utilities included in the Distribution System Cost Methodologies paper by Shirley W.
(2001) for the Regulatory Policy Project's Distributed Resource Policy Series.
Reduction in Wholesale Market Clearing Prices
E3 and RMI, Methodology and Forecast of Long Term Avoided Costs for the Evaluation
of California Energy Efficiency Programs, October 26, 2004
Hadley, S.W., et al. 2003. Quantitative Assessment of Distributed Energy Resource
Benefits, (page 8-19)
Heschong Mahone Group, Inc., Ridge & Associates, Energy and Environmental
Economics, Inc. 2005. New York Energy Smart Program Cost-Effectiveness
Assessment June 2005. Prepared for: The New York State Energy Research and
Development Authority (NYSERDA). (page 23 and 39)
Synapse Energy Economics. 2006. Portfolio Management: Tools and Practices for
Regulators, prepared for the national Association of Regulatory Utility Commissioners
(NARUC), July 17, 2006.
Increased Reliability and Power Quality
GE Corporate Research and Development. 2003. DG Power Quality, Protection, and
Reliability Case Studies Report. Prepared for NREL August 2003
IEEE Std. 1366-1998: Trial Use Guide for Electric Power Distribution Reliability Indices.
Organization: IEEE
New Power Technologies presented a methodology to evaluate the impacts of
distributed energy resources on the grid operation in its 2005 paper "Optimal
Portfolio Methodology for Assessing Distributed Energy Resources Benefits for the
ENERGYNETSM" prepared for California Energy Commission. April 2005.
ORNL TM-2004/91, Measurement Practices for Reliability and Power Quality: A
Toolkit of Reliability Measurement Practices, John D. Kueck, Brendan J. Kirby, Philip N.
Overholt , Lawrence C. Markel , June 2004.
Avoided Risks
http://www.pjm.com/index.jsp
http://www.synapse-energy.com/Downloads/
SynapseReport.2006-07.NARUC.Portfolio-
Management-Tools-and-Practices-for-
Regulators.05-042.pdf
http://www.sec.gov/edgar/searchedgar/
companysearch.html
h ttp://www.naruc. org/
http://www.masstech.org/dg/Benefits.htm
http://www.raponline.org/Pubs/DRSeries/
CostTabl.zip
http://www.ethree.com/CPUC/E3_Avoided_
Costs_Final.pdf
http://www.ornl.gov/~webworks/cppr/y2001/
rpt/116227.pdf
http://www.nyserda.org/Energy_lnformation/
ContractorReports/Cost-Effectiveness_Report_
June05.pdf
http://www.synapse-energy.com/Downloads/
SynapseReport.2006-07.NARUC.Portfolio-
Management-Tools-and-Practices-for-
Regulators.05-042.pdf
http://www.localpower.org/documents/
reporto_nrel_powerquality.pdf
http://standards.ieee.org/reading/ieee/std_
public/description/td/1366-1998_desc.html
http://www.energy.ca.gov/2005publications/
CEC-500-2005-096/CEC-500-2005-096.PDF
http://www.ornl.gov/sci/btc/apps/
Restructuring/ORNLTM200491FINAL.pdf
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 88
-------
Biewald, Bruce, Tim Woolf, Amy Roschelle, and William Steinhurst 2003. Portfolio
Management: How to Procure Electricity Resources to Provide Reliable, Low-Cost,
and Efficient Electricity Services to All Retail Customers. Synapse Energy Economics.
Prepared for the Regulatory Assistance Project and the Energy Foundation. October,
2003.
Brealey, Richard A., and Stewart C. Myers 2002. Principles of Corporate Finance, Irwin
McGraw-Hill, New York.
Dixit, A.K. & R.S. Pindyck 1994. "Investment under Uncertainty." Princeton University
Press, Princeton, N.J., 1994.
Lovins, A.B., E. K. Datta, T. Feiler, K. R. Rabago, J.N. Swisher, A. Lehmann, and K.
Wicker 2002. Small is Profitable: the Hidden Economic Benefits of Making Electrical
Resources the Right Size.
New England Electric System 1993. NEESPLAN 4. Westborough, MA: New England
Electric System, Nov. 1993. California ISO Midwest ISO
"The Professional Risk Managers' Guide to Energy & Environmental Markets," edited
by Peter C. Fusaro, PRMIA, 2006.
Fuel and Technology Diversification
EIA Form EIA-906 and 920 (power plant database) - now EIA-923
Grubb, M., Butler, L, and Twomey, P., 2005. Diversity and security in UK electricity
generation: The influence of low-carbon objectives. Energy Policy, article in press.
Portfolio management tools, programs, and skill sets for utilities and utility
commissions ("Portfolio Management: Tools and Practices for Regulators" Synapse,
2006)
http://www.raponline.org/Pubs/
PortfolioManagement/SynapsePMpaper.pdf
N/A
N/A
h ttp://www.smallisprofitable. org/
N/A
N/A
http://www.eia.doe.gov/cneaf/electricity/page/
eia906_920.html
http://dx.doi.0rg/10.1016/j.enpol.2005.09.004
http://www.synapse-energy.com/Downloads/
SynapseReport.2006-07.NARUC.Portfolio-
Management-Tools-and-Practices-for-
Regulators.05-042.pdf
AESC 2005. Avoided-Energy-Supply-Component (AESC) Study Group. Avoided Energy
Supply Costs in New England. December 23 2005.
Barbose, G., Ryan Wiser, Amil Phadke and Charles Goldman. "Reading the Tea Leaves:
How Utilities in the West are managing Carbon Regulatory Risk in their Resource
Plans." LBNL, March 2008.
Baskette, C., B. Horii, E. Kollman, S. Price. 2006. "Avoided Cost Estimation and Post-
reform Funding Allocation for California's Energy Efficiency Programs." In Energy.
Elsevier. Vol. 31, pp. 1084-1099.
Biewald, B. 2005. Using Electric System Operating Margins and Build Margins in
Quantification of Carbon Emission Reductions Attributable to Grid Connected COM
Projects. Prepared for the UN FCCC. September 19, 2005.
Boston Globe, "Marshfield to be test site for new energy-saving plan," March 30,
2008.
Burkhart, Lori A. "FERC Takes on Reactive Power," Fortnightly's SPARK, March 2005.
http://publicservice.vermont.gov/pub/
aescstudy.html
http://eetd.lbl.gov/ea/emp/reports/lbnl-44e.pdf
N/A
http://cdm.unfccc.int/Panels/meth/meeting/05/
Methl7_repanl2_BiewaldPaperOMBMMargins.
pdf
h ttp://www. bos ton.com/news/local/
articles/2008/03/30/marshfield_ to_be_ test_
site_for_new_energy_saving_plan/
http://www.pur.com/pubs/spark/mar05.pdf
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 89
-------
References URL Address
California Public Utilities Commission (CPUC) 2006. Interim Opinion: 2006 Update of
Avoided Costs and Related Issues Pertaining to Energy Efficiency Resources. Decision
06-06-063. June 29.
California PUC 2004. Interim Opinion: Energy Savings Goals for Program Year 2006
and Beyond. Decision 04-09-060, September 23, 2004.
CA PUC 2004. Decision 04-12-048, December 16, 2004.
California Public Utilities Commission (CPUC) 2005. Interim Opinion: Energy
Efficiency Portfolio Plans and Program Funding Levels for 2006-2008 - Phase 1
Issues. Decision 05-09-043. September 22, 2005.
Energy and Environmental Analysis, Inc. 2003. Natural Gas Impacts of Increased CHP.
Prepared for U.S. Combined Heat and Power Association. October 2003.
Energy and Environmental Economics, Inc. (E3) and Rocky Mountain Institute (RMI).
2004. Methodology and Forecast of Long Term Avoided Costs for the Evaluation
of California Energy Efficiency Programs. Prepared for California Public Utilities
Commission Energy Division. October 25.
EIA Electric Power Monthly, Table 4.13.A: Average Cost of Natural Gas Delivered for
Electricity Generation by State.
Feinstein, C.D., R. Organs, and S.W. Chapel 1997. The Distributed Utility: A New
Electric Utility Planning and Pricing Paradigm. Annual Review of Energy and
Environment. 22. 155-185.
Heschong Mahone Group, Inc. 2005. New York Energy Smart Program Cost
Effectiveness Assessment. Prepared for the New York State Energy Research and
Development Authority. June.
ISO NE, Monthly Summary of Hourly Data (2006)
Jensen, V., and E. Lounsbury. 2005. Assessment of Energy Efficiency Potential
in Georgia. Prepared for the Georgia Environmental Facilities Authority by ICF
Consulting. May.
Johnston, L, A. Roschelle, and B. Biewald, "Taking Climate Change into Account in
Utility Planning: Zero is the Wrong Carbon Value," Synapse Energy Economics, March
17, 2005.
Kee. 2005. Testimony of Edward D. Kee before the Public Utilities Commission of
South Dakota, EL04-16. January 31.
LaCapra Associates. 2003. Alternatives To Velco's Northwest Vermont Reliability
Project. January 29.
Lawrence Berkeley National Laboratory 2002. California Customer Load Reductions
during the Electricity Crisis: Did they Help to Keep the Lights On? May 2002.
http://www.cpuc.ca.gov/PUBLISHED/
COMMENT_DECISION/56572.htm#P86_2251
http://www.cpuc.ca.gov/PUBLISHED/FINAL_
DECISION/40212.htm
h ttp://www. cpuc. ca.gov/WORD_ PDF/FIN AL_
DECISION/43224.PDF
http://www.cpuc.ca.gov/PUBLISHED/FINAL_
DECISION/49859.htm
h ttp://www. eea - in c. com/dgchp_ presen ta tions/
CHP-Gas-Finalppt
http://ethree.com/CPUC/E3_Avoided_Costs_
Finalpdf
See also updates to the initial report:
http://ethree.com/cpuc_avoidedcosts.html
http://www.eia.doe.gov/cneaf/electricity/epm/
table4_13_a.html
http://arjournals.annualreviews.
org/doi/abs/10.1146/annurev.
energy.22.1.155?journalCode = energy
www.nyserda.org/EnergyJnformation/
ContractorReports/Cost-Effectiveness_Report_
June05.pdf
http://www.iso-ne.com/markets/hstdata/index.
html
http://www.gefa.org/Modules/ShowDocument
aspx?documentid=46
http://www.synapse-energy.com/Downloads/
SynapseReport 2005- 03. CA CI-HEC.Taking-
Climate-Change-into-Account--Zero-is-the-
Wrong-Value.04-44.pdf
http://www.synapse-energy.com/Downloads/
SynapseTestimony.2005-02.SD-PSC-.Avoided-
Costs-for-the-Java-Wind-Project.04-74.pdf
http://207.136.225.66/Downloads/Other/
VELCO.pdf
http://repositories.cdlib.org/cgi/viewcontent
cgi?article=15966-context=lbnl
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 90
-------
URL Address
Lovins, A.B., E. K. Datta, T. Feiler, K. R. Rabago, J.N. Swisher, A. Lehmann, and K.
Wicker 2002. Small is Profitable: the Hidden Economic Benefits of Making Electrical
Resources the Right Size.
Massachusetts Technology Collaborative (MTC) 2006 (a). Attachment H - Report of
Distribution Planning Work Group on DG and Distribution Deferral
Massachusetts Technology Collaborative (MTC) 2006. Massachusetts Distributed
Generation Collaborative 2006 Report under D.T.E. 02-38-C Investigation by the
Department of Telecommunications and Energy on its own Motion into Distributed
Generation. June 30, 2006.
NARUC (National Association of Regulatory Utility Commissioners) 1992. Electric
Utility Cost Allocation Manual, January 1992.
Navigant Consulting Inc. (NCI) 2006. Distributed Generation and Distribution
Planning: An Economic Analysis for Massachusetts DG Collaborative. January 20,
2006. Prepared for the Massachusetts Technology Collaborative
NERC. Glossary of Terms.
NRDC (Natural Resources Defense Council) Chang, Audrey. California's Sustainable
Energy Policies Provide A Model for the Nation. March 2006.
Orans, R., Area-Specific Marginal Costing for Electric Utilities: A Case Study of
Transmission and Distribution Costs, 1989. PhD Thesis, Stanford University.
Orans, R., C.K. Woo, J.N. Swisher, B. Wiersma and B. Horii, 1992. Targeting DSM for
Transmission and Distribution benefits: A Case Study of PG&E's Delta District, Report
No. TR-100487, Electric Power Research Institute.
Orans R., S. Price, D. Lloyd, T. Foley, E. Hirst 2001. Expansion of BPA Transmission
Planning Capabilities. November 2001. Prepared for Transmission Business Line
Bonneville Power Administration
Pace Energy and Synapse 2006. A Comprehensive Process Evaluation of Early
Experience Under New York's Pilot Program for Integration of Distributed Generation
in Utility System Planning. Prepared for The New York State Energy Research and
Development Authority, (forthcoming)
Pacific Northwest National Laboratory. 2004. Avoiding Distribution System Upgrade
Costs Using Distributed Generation
PJM LLC, RTO. 2006(a). "PJM Region Sets Summer's Third Energy-Use Record: Energy
Conservation Request Continues East." August 2, 2006.
PJM LLC, RTO. 2006(b). "Early Aug. Demand Response Produces $650 Million Savings
in PJM: Reducing Electricity Use Stretches Power Supplies, Lowers Wholesale
Electricity Supplies. "August 17, 2006.
Roschelle, A. and W. Steinhurst. (2004) "Best Practices in Procurement of Default
Electric Service: A Portfolio Management Approach," Electricity Journal, Oct. 2004
Shirley W. 2001. For the Regulatory Assistance Project (RAP.) Distribution System Cost
Methodologies for Distributed Generation. September 2001.
Stoddard, L, J. Abiecunas, and R. O'Connell. 2006. Economic, Energy, and
Environmental Benefits of Concentrating Solar Power in California. Prepared by Black
& Veatch for U.S. DOE National Renewable Energy Laboratory. April.
http://www.smallisprofitable.org/
h ttp://www. mtpc. org/dg/collab- reports, h tm
h ttp://mass tech. org/dg/02-38- C_
Attachment-H_DPWG_Report.pdf
http://www.naruc.org/Store/
http://www.masstech.org/renewableenergy/
public_policy/DG/2006-01-23_MTC_Navigant_
DG-Economic-Analysis.pdf
h ftp://www.nerc. com/files/Glossary_l 2Feb08. pdf
http://docs.nrdc.org/air/air_06033101a.pdf
N/A
N/A
http://www.transmission.bpa.gov/PlanProj/
bpa_ tbl_planning.pdf
http://www.synapse-energy.com/Downloads/
SynapseReport.2006-08.NYSERDA.DG-in-New-
York-with-Pace-Law-School.04-37.pdf
http://gridwise.pnl.gov/docs/pnnlsa39886.pdf
http://www.pjm.eom/about-pjm/newsroom/~/
media/about-pjm/newsroom/2006-releases/
20060802-pjm-sets-third-peak-record.ashx
http://www.pjm.eom/about-pjm/newsroom/~/
media/about-pjm/newsroom/2006-releases/
20060817-demand-response-savings.ashx
N/A
http://www.raponline.org/Pubs/DRSeries/
DistCost.PDF
http://www.nrel.gov/docs/fy06osti/39291.pdf
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 91
-------
Synapse Energy Economics 2003. Modeling Demand Response and Air Emissions in
New England. August 2003 .
URL Address
http://www.synapse-energy.com/Downloads/
SynapseReport.2003-09.US-EPA.NE-DR-and-
AE-Modeling.03-01.pdf
Synapse Energy Economics. 2006. Portfolio Management: Tools and Practices for
Regulators, prepared for the national Association of Regulatory Utility Commissioners
(NARUC), July 17, 2006.
http://www.synapse-energy.com/Downloads/
SynapseReport.2006-07.NARUC.Portfolio-
Management-Tools-and-Practices-for-
Regulators.05-042.pdf
Tellus Institute 1995. "Costing Energy Resource Options: An Avoided Cost handbook
for Electric Utilities," for US EPA and US DOE, September 1995.Shirley W. 2001. For
the Regulatory Assistance Project (RAP.) Distribution System Cost Methodologies for
Distributed Generation. September 2001.
http://telluspublications.stores.yahoo.net/
energygroup.h tml
CHAPTERS | Assessing the Multiple Benefits of Clean Energy 92
-------
CHAPTER FOUR
Assessing the Air Pollution, Greenhouse
Gas, Air Quality, and Health Benefits of
Clean Energy Initiatives
Many states and localities are
exploring or implementing
clean energy policies to achieve
greenhouse gas (GHG) and criteria
air pollutant1 emission reductions.
For example, New Mexico's Climate Change Advisory
Group Action Plan estimates that clean energy mea-
sures could achieve more than one-third of the 35
million metric tons of potential carbon dioxide (CO2)
reductions identified in New Mexico in 2020, repre-
senting around 15 percent of the projected baseline
emissions levels in 2020 (New Mexico Climate Change
Advisory Group, 2006). The Metropolitan Washington
Council of Governments included renewable energy
and energy efficiency measures in its May 2007 State
Implementation Plan (SIP) for the 8-Hour Ozone
Standard. These measures are expected to avoid almost
150,000 MWh of generation and 0.17 tons of NOx daily
(Metropolitan Washington COG, 2007).
GHG and criteria air pollutant emission reduction
estimates are important measures of the potential or
realized benefits of clean energy, and are a critical first
step for further environmental benefits analysis. Once
emitted, some criteria air pollutants are transported in
1 Criteria air pollutants are particle pollution (often referred to asp articulate
matter), ground-level ozone, carbon monoxide, sulfur oxides, nitrogen oxides,
and lead. The Clean Air Act requires EPA to set National Ambient Air Quality
Standards for these air pollutants. EPA calls these pollutants "criteria" air
pollutants because it regulates them by developing human health-based and/
or environmentally based criteria (i.e., science-based guidelines) for setting
permissible levels (U.S. EPA, 20084).
U
o
Q
CHAPTER ONE
Introduction
CHAPTER TWO
Potential Energy Impacts of Clean Energy
CHAPTER THREE
Electric System Benefits of Clean Energy
CHAPTER FOUR
Air Quality Benefits of Clean Energy
CHAPTER FIVE
Economic Benefits of Clean Energy
APPENDIX A
Catalogue of Clean Energy Case Studies
APPENDIX B
Tools and Models Referenced in Each Chapter
CHAPTER FOUR CONTENTS
4.1 How Clean Energy Initiatives Result in Air and
Health Benefits
95
4.2 How States Estimate the GHG, Air, and Health Benefits
of Clean Energy 98
4.2.1 Step 1: Develop and Project a Baseline
Emissions Profile 98
4.2.2 Step 2: Quantify Air and GHG Emission
Reductions from Clean Energy Measures 107
4.2.3 Step 3: Quantify Air Quality Impacts 115
4.2.4 Step 4: Quantify Human Health and Related
Economic Effects of Air Quality Impacts 119
4.3 Case Studies 124
4.3.1 Texas Emissions Reduction Plan (TERP) 124
4.3.2 Wisconsin - Focus on Energy Program 125
Information Resources 127
References 130
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 93
-------
the atmosphere potentially for long distances. Some
"primary" pollutants are directly harmful to exposed
humans and the environment, while other "secondary"
pollutants can affect human health after they form as
a result of photochemical reactions in the atmosphere.
For example, nitrogen oxides (NOX) and volatile organic
compounds (VOCs) react under certain meteorological
conditions to form ozone (O3), a principal component
of photochemical smog. Estimating the impact of
changes in criteria air pollutant emissions on ambient
air quality and the related environmental and health im-
pacts can enhance a state's understanding of the poten-
tial benefits that can result from clean energy measures.2
Understanding a range of environmental and human
health benefits from existing and proposed clean en-
ergy measures can help state planners:
1. Identify opportunities where meeting today's
energy challenges can serve as an environmental
improvement strategy,
2. Potentially reduce the compliance costs of meeting
air quality standards by offering more options to
states, and
3. Build support for clean energy initiatives among
state and local decision makers.
This chapter is designed to help states understand the
methods, models, opportunities, and issues associated
with assessing the GHG, air pollution, air quality, and
human health benefits of clean energy options. While
it focuses primarily on emissions from electricity, the
methods and tools presented in this chapter could be
applied to emissions from other sources.
Section 4.1, How Clean Energy Initiatives Result
in Air and Health Benefits, describes the environ-
mental and health benefits of clean energy and
addresses several key issues associated with esti-
mating these benefits.
Section 4.2, How States Estimate the GHG, Air, and
Health Benefits of Clean Energy, presents four key
steps a state can take to estimate the air and health
2 By influencing climate change, GHGs can indirectly lead to air quality
and health effects. Climate change can lead to more frequent extreme heat
events and exacerbate air quality problems through increased temperatures.
Methane, which is a key GHG, contributes to ground-level ozone formation.
Criteria air pollutants, however, are directly linked to changes in air quality
and human health effects in scientific literature. For this reason, this chapter
addresses the air quality and human health benefits associated with reducing
criteria air pollutant, but not GHG, emissions.
STATES ARE QUANTIFYING THE ENVIRONMENTAL BENEFITS
OF CLEAN ENERGY POLICIES
Several states have quantified the emission reductions and
air and health benefits from their clean energy measures and
determined that the measures are helping them reduce their air
pollution and GHGs.
A recent evaluation of The Wisconsin Focus on Energy
Program's energy efficiency and renewable energy projects
funded by the Utility Public Benefits fund, for example,
shows that during the period from program inception in July
2001 through June 30, 2006, the state has displaced annual
emissions from power plants and utility customers of about:a
5.8 million pounds of NOX,
2.6 billion pounds of CO2,
11.4 million pounds of SO2, and
46 pounds of mercury (Hg)
In 2004, the Texas Commission on Environmental Quality
evaluated the Texas Emissions Reduction Plan and calculated
that it achieves an annual reduction of NOx emissions of 346
tons through energy efficiency and renewable energy. NOx
reductions over the period 2007-2012 are projected to range
from 824 tons per year in 2007 to 1,416 tons per year in 2012.
Sources: DOA, 2006; Haberl et al, 2007
a These emission values vary greatly by type of pollutant, due
primarily to the content of carbon, sulfur, nitrogen, and mercury
in fossil fuels. For example, CO2 emission reductions from clean
energy programs are comparatively high because fossil fuels
are rich in carbon, and CO2 is a primary product of fossil fuel
combustion. On the other hand, the concentration of Hg in fuel
(primarily coal) is very small, and so emission reductions of Hg
are also small compared with reductions of other pollutants.
benefits of clean energy and describes related
methods, tools and issues.
Section 4.2.1, Step 1: Develop and Project a
Baseline Emissions Profile, focuses on develop-
ing and projecting an emissions inventory to
establish a baseline from which progress can
be measured.
> Section 4.2.2, Step 2; Quantify Air and GHG
Emission Reductions from Clean Energy Mea-
sures, provides guidance on quantifying GHG
and criteria air pollutant emission reductions
that result from clean energy measures.
Section 4.2.3, Step 3: Quantify Air Quality
Impacts, describes how to estimate the changes
in air quality that result from air pollution
emission reductions.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 94
-------
Section 4.2.4, Step 4: Quantify Human Health
and Related Economic Effects of Air Quality
Impacts, addresses the quantification of public
health impacts based on estimates of air pollu-
tion or air quality changes.
Section 4.3, Case Studies, presents examples of how
two states, Texas and Wisconsin, have estimated
the air quality and health benefits resulting from
their clean energy programs.
4.1 HOW CLEAN ENERGY INITIATIVES
RESULT IN AIR AND HEALTH BENEFITS
Electricity generation from fossil fuels is a major
source of many types of air pollution, including GHGs
and criteria air pollutants. These emissions contribute
to a variety of environmental issues, including global
warming and human health problems, which are de-
scribed below.
GHG emissions occur naturally and absorb some of
the heat that would otherwise escape to space (see
Figure 4.1.1, The Greenhouse Effect). GHGs keep the
planet warmer than it would otherwise be through this
natural "greenhouse effect." Human activity-related, or
anthropogenic, GHGs, such as those from electricity
generation, are increasing the greenhouse effect and
are very likely responsible for most of the observed
increase in global average temperatures since the mid-
20th century.
The process of generating electricity from fossil fuels is
the single largest source of anthropogenic carbon diox-
ide (CO2) emissions in the United States, representing
39 percent of CO2 emissions in 2006 (U.S. EPA, 2008b).
GHGs are also emitted during the refinement, process-
ing, and transport of fossil fuels. These gases accumu-
late and can remain in the atmosphere for decades to
centuries, affecting the global climate system for the
long term. Measures to reduce GHGs in the near term,
therefore, may have a large impact on our ability to
meet long-term climate objectives.
Criteria air pollutants affect air quality and human
health directly and in the short term. The use of fossil
fuels for electric generation causes increased levels of
these pollutants in the atmosphere. Some criteria pol-
lutants, including particle pollution (often referred to
as particulate matter or PM), carbon monoxide, sulfur
dioxide (SO2) and nitrogen oxides (NOx), are directly
emitted into the atmosphere as the result of fossil fuel
combustion. Ozone (O3) and fine particulate matter
FIGURE 4.1.1 THE GREENHOUSE EFFECT
The Greenhouse Effect
Some solar radiation
is reflected by
the earth and the
atmosphere
Some of the infrared radiation passes through the
atmosphere. Some is absorbed and re-emitted in
all directions by greenhouse gas molecules. The
effect of this is to warm the earth's surface and
the lower atmosphere.
Most radiation is absorbed
by the earth's surface and
Atmosphere
th's Surface
Infrared radiation
is emitted by the
earth's surface
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 95
-------
(PM25) are "secondary" pollutants that form in the air
when directly emitted criteria pollutants and other
precursor air pollutants, such as volatile organic com-
pounds (VOCs), react or interact. O3 and PM25 are of
particular concern because they are most prevalent and
are linked with a variety of respiratory and cardiovas-
cular illnesses and death.3
GHGs and criteria air pollutants have different effects
on air quality and human health due to their different
temporal and spatial characteristics. While GHGs have
a global effect and can last more than 100 years, criteria
air pollutants have a local to regional effect on air
quality and human health, and can dissipate in hours
or days. Clean energy measures that reduce criteria air
pollutants, therefore, can result in almost immediate
local improvements in air quality and human health. In
addition, the location and timing of the emissions from
criteria air pollutants is very important in determining
how significantly they affect human health. Since these
pollutants tend to dissipate over time and space, those
that occur far away from populations will have less of
an impact on human health than those closer to dense-
ly populated areas. In contrast, the impact of GHGs on
the overall climate system is not affected by the specific
location of an emission. One ton of GHG emitted in
one location affects the global climate system the same
as one ton of the same GHG in a different location.
Clean energy measures reduce the emission of the pol-
lutants described above and related effects on health or
the global climate by reducing demand for fossil fuel-
based electricity through either:
Reducing total electric demand through energy
efficiency, or
Directly displacing conventional electricity sup-
plies with clean distributed generation (DG) or
renewable energy sources.
The impact of any kind of clean energy resource on
air pollutant and GHG emissions and its subsequent
effect on human health or global climate change varies
depending on the generation sources that are displaced
and the resource that is displacing the generation.4
3 Tropospheric O} also acts as a strong GHG. Different components ofPM25
have both cooling (e.g., sulfates) and warming (e.g., black carbon) effects on
the climate system.
4 DG and combined heat and power (CHP) units often burn fossil fuels as
their primary fuel source. In this case, net emissions (i.e., displaced emissions
less the emissions of the DG or CHP unit) also depend on the technology and
fuel source for the DG or CHP unit.
FIGURE 4.1.2 NERC INTERCONNECTIONS
NERC INTERCONNECTIONS
QUEBEC
INTERCONNECTION
WESTERN
INTERCONNECTION
EASTERN
INTERCONNECTION
ERCOT
NTERCONNECTION
Source: NERC, 2008.
To estimate emission reductions associated with clean
energy, it is important to determine which resources
are expected to be displaced. This was discussed in de-
tail in Chapter 3 and is repeated here in summary form
for completeness. Estimating the emissions associated
with the displaced generation presents challenges due
to (1) the complex way that electricity is generated
and transmitted across the United States and (2) un-
certainty about the future location of emissions due to
market-based environmental programs such as cap and
trade. These challenges are discussed below.
Electricity Generation, Transmission, and Distri-
bution. The continental United States and Canada
are divided into four interconnected alternating
current (AC) grids (the Eastern, Western, Quebec,
and Electric Reliability Council of Texas [ERCOT]
Interconnections) as depicted in Figure 4.1.2,
NERC Interconnections. Each of the grids is electri-
cally isolated with only a limited number of direct
current (DC) ties connecting them. However,
within each of these grids, electricity is imported
or exported continuously among the numerous
power control areas.
The demand for electricity varies by season and by
time of day. Some power plantsthe baseload coal
plants and other plants with low variable operating
costs such as nuclear and hydroelectricoperate
at very high levels. The output of other generators
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 96
-------
CLEAN ENERGY AND LEAKAGE
The goal of clean energy policies is typically to reduce emissions
within the state or larger region where the policies are
implemented. However, due to the interconnected and dynamic
nature of the power system, the benefits of clean energy
policies may not be completely realized within the region. As
utilities and control area operators seek to operate the system
to minimize the cost of providing electricity, power transactions
occur across the area, both on a long-term contract basis and
on a spot basis. As a result, reductions in electricity demand in
the region where clean energy policies have been implemented
may not always result in corresponding reductions in electricity
generation in the same region, depending on the relative cost of
this generation and that of neighboring regions.
Reductions in electricity demand levels in the Mid Atlantic region
from clean energy policies, for example, might be expected to
reduce generation in the Mid Atlantic region. However, if the
cost of this now-excess generation in the Mid Atlantic is less
than the neighboring regions' marginal sources of generation,
it may be economic to use these now-available resources to
meet demand in those neighboring regions, thereby displacing
more expensive generation. For example, clean energy policies
put in place in Pennsylvania may result in reduced emissions in
the New England region as lower cost, coal-fired generation is
freed up to displace more expensive oil- or gas-fired steam units
in New England. The extent of these generation and associated
emissions shifts will depend on the cost differential, available
transmission capacity, reliability considerations, environmental
constraints, and a number of other factors. This shifting of
displaced resources from one area to another is often called
"leakage" and is an important consideration when assessing the
emissions benefits of clean energy programs.
rises and falls throughout the day, responding to
changing electricity demand. Other generators are
used as "peaking" units and are operated only dur-
ing the times of highest demand.
A group of system operators across the region de-
cides when and how to dispatch electric generation
from each power plant in response to the demand
at the time. System operators decide which power
plants to dispatch and in what order based on
demand at that moment and the cost or bid price.
Baseload plants are dispatched first. These plants
are typically characterized as having low operat-
ing costs, and may be operated at a constant rate.
Examples include coal and nuclear plants. Peaking
units are dispatched last. These units are typically
characterized as having high operating costs, and
also have the ability to be dispatched quickly. Exam-
ples include natural gas turbines and diesel genera-
tors. The fuels, generation efficiencies, control tech-
nologies, and emission rates vary greatly from plant
to plant by season and time of day. The emissions
effects of energy demand reductions, therefore,
also vary by load levels, time of day, and season. As
discussed later in this section, the interconnected
basis of the system, along with least-cost dispatch
practices, has implications for the impacts of the
effectiveness of clean energy programs in the region
in which they are implemented. Specifically, there is
potential for generation and emissions leakage from
the implementing region to neighboring regions if
specific measures are not taken to limit this.
Other conditions besides demand and cost affect
dispatch. Transmission constraints, when transmis-
sion lines become congested, can make it difficult
to dispatch power from far away into areas of high
electricity demand. Extreme weather events can
decrease the ability to import or export power from
neighboring areas. "Forced outages," when certain
generators are temporarily not available, can also
shift dispatch to other generators. System operators
must keep all these issues in mind when dispatch-
ing power plants. For more information about how
the electric system works, see Section 3.1, How
Clean Energy Can Achieve Electric System Benefits.
Air Emission Cap and Trade Programs. Air
emission cap and trade programs, such as the Acid
Rain Program,5 set annual limits (i.e., caps) on
fossil-fuel-fired electric generators' emissions and
play an important role in ensuring that air pollut-
ant emissions are reduced.
Under cap and trade programs, each utility or
generator typically receives a certain number of
allowances, each of which is an authorization to
emit one ton of a specific air pollutant (e.g., SO2). A
generator must obtain enough allowances to cover
its emissions. If a generator has excess allowances,
due, for example, to the installation of air pollution
control devices, it can bank the allowances for later
use or sell the allowances to another company, de-
pending upon the specific program rules. If a gen-
erator does not expect to have enough allowances
to authorize its emissions, it can buy allowances,
install emissions controls, or curtail its activity.
The trading component of the cap and trade
program allows for the most cost-effective emis-
sion reductions to occur first. If the demand for
5 The Acid Rain Program regulates SO2 and NO emissions in the continental
United States to reduce acid deposition caused by these emissions.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 97
-------
allowances decreases or the supply of excess allow-
ances increases (e.g., because clean energy mea-
sures result in reduced fossil-fuel-fired electricity
generation) the cost of achieving the cap decreases,
but the cap itself does not change. While cap and
trade programs ensure a certain reduced level
of emissions and can result in a more diversified
energy system, trading emission allowances means
that it can be difficult to attribute emission reduc-
tions to specific clean energy measures, and that in
some cases clean energy measures may not result
in net emission reductions at all.
Despite these challenges, tools and methods exist for
states to address these issues and estimate air emission
reductions, air quality changes, and human health
effects associated with clean energy policies. These
approaches are described below in Section 4.2, How
States Estimate the GHG, Air and Health Benefits of
Clean Energy.
4.2 HOW STATES ESTIMATE THE
GHG, AIR, AND HEALTH BENEFITS OF
CLEAN ENERGY
Analysis to quantify the greenhouse gas, air pollution,
air quality, and human health benefits of clean energy
initiatives involves four basic steps:
1. Develop and project a baseline emissions inventory,
2. Quantify the air and GHG emission reductions
from the clean energy measures,
3. Estimate the changes in air quality resulting from
these emission reductions, and
4. Estimate the human health and related economic
effects of these air quality changes.
These steps often occur linearly, as shown in Table 4.2.1,
Steps for Estimating GHG, Air, and Health Benefits of
Clean Energy Initiatives. This is because estimating some
of the benefits, such as improved air quality and reduced
human health effects, requires information generated
in previous stepsspecifically the timing and type of
generation displaced by the clean energy measures.
Some states may not be interested in estimating all of
the benefits described in this section, or they may not
achieve benefits in each area. For example, as described
in Section 4.1, How Clean Energy Initiatives Result in
Air and Health Benefits, while criteria air pollutants
GUIDANCE ON CREDITS FOR EMISSION REDUCTIONS
FROM CLEAN ENERGY
EPA has developed a State Implementation Plan (SIP) guidance
document that provides a step-by-step procedure for
quantifying the benefits. It describes the following two options
for state and local governments to address the presence of a
cap and trade program when quantifying emission reductions
from clean energy:
Retire commensurate amount of allowances, or
Demonstrate that an emission or air quality benefit is
expected to occur even in the presence of such a cap and
trade program.
Source: U.S. EPA, 2004.
are linked directly with air quality changes and human
health effects, greenhouse gas emissions are indirectly
linked to air quality and human health effects.6 Thus,
if a state clean energy policy yields GHG impacts but
very low criteria air pollutant impacts, it may not be
worthwhile to continue evaluating the air quality and
subsequent health impacts because they likely would
be negligible.
The remainder of this section describes basic and so-
phisticated modeling approaches, and related protocols,
data needs, tools, and resources that states can use dur-
ing each step in the process of quantifying the GHG, air,
and human health benefits of clean energy initiatives.
4.2.1 STEP 1: DEVELOP AND PROJECT
A BASELINE EMISSIONS PROFILE
The initial step in measuring clean energy emissions
reductions is to prepare a state-level emissions inven-
tory and projection that documents the baseline, or
the emissions that occur without any additional clean
energy policies. This baseline can include historical,
current, and projected emissions data and provides a
clear reference case against which to measure the emis-
sion impacts of a clean energy initiative.
Emissions inventories and projections are typically
created for criteria air pollutants to support air quality
attainment planning, or for GHGs to support climate
change action plans, but do not necessarily include
both GHGs and criteria air pollutants. However, an
inventory that includes both types of emissions will
6 Nevertheless, clean energy measures that reduce GHGs may also reduce
criteria air pollutants, thus resulting in direct health benefits.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 98
-------
TABLE 4.2.1 STEPS FOR ESTIMATING GHG, AIR, AND HEALTH BENEFITS OF CLEAN ENERGY INITIATIVES
>pand
e Emissio
(Section 4.2.1)
a. Select method.
b. Compile criteria air
pollutants from available
sources into inventory.
c. Develop a forecast using
assumptions about future and
available tools.
Greenhouse Gas Emissions
a. Select method.
b. Compile greenhouse gas
emissions from available
sources into inventory.
c. Develop a forecast using
assumptions about future and
available tools.
Clean Energy Measures
(Section 4.2.2)
(Section 4.2.3)
Air Quality Impacts
(Section 4.2.4)
a. Develop criteria air pollutant
reductions from clean energy
using:
energy savings estimates,
operating characteristics of
clean energy resource (load
profile),
emissions factors, and
control technology data.
Compare against the baseline.
Use criteria air pollutant data
to estimate changes in air
quality with an air quality
model.
a. Use data on air quality
changes and epidemiological
and population information to
estimate health effects.
b. Apply economic values
of avoided health effects to
monetize benefits.
a. Develop greenhouse gas
emission reductions from
clean energy using:
energy savings estimates
and a profile of when these
impacts will occur,
operating characteristics of
clean energy resource,
emissions factors, and
fuel data.
b. Compare against the
baseline.
n/a
n/a
facilitate a more comprehensive analysis of the emis-
sions benefits of clean energy and the value of clean
energy policies. This is important because many op-
tions that reduce GHGs may, in fact, reduce criteria air
pollutants and indirectly yield health benefits. On the
other hand, some measures that reduce GHG emissions
can actually increase emissions of criteria air pollutants.
For example, a measure that encourages switching
from electricity generated with natural gas to electric-
ity generated by wind will result in both criteria air
pollutant benefits and GHG emission reductions. The
impact on air pollution is less certain, however, if a
state switches from natural gas to biomass-generated
energy. It is important to take these considerations into
account when evaluating the air and health benefits
of clean energy measures. Developing a baseline that
includes both GHGs and criteria air pollutants serves
as a future point of reference for retrospective program
evaluation as well as a basis for making well-informed
policy and planning decisions.
Typically, a state's air agency creates the criteria air
pollutant inventory every three years as part of its
responsibility to meet National Ambient Air Quality
Standards established under the Clean Air Act. GHG
emissions inventories can be developed by state air
or other agencies, but since states are not required
by federal law to inventory their GHG emissions, the
practice varies from state to state. State energy offices
or universities sometimes develop GHG inventories
on an annual basis or every few years. If inventories
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 99
-------
SOURCES OF AIR POLLUTION EMISSIONS
Air emission sources are grouped into four categories: point,
area, mobile (on-road and non-road), and biogenic sources.
Each is described below.
Point Source: A stationary location or fixed facility from which
pollutants are discharged, such as an electric power plant or a
factory smokestack.
Area Source: An air pollution source that is released over a
relatively small area but cannot be classified as a point source.
Area sources include small businesses and household activities,
product storage and transport distribution (e.g., gasoline),
light industrial/commercial sources, agriculture sources (e.g.,
feedlots, crop burning), and waste management sources (e.g.,
landfills). Emissions from area sources are generally reported by
categories rather than by individual source.
On-Road Mobile Source: Sources of air pollution from highway
vehicles such as cars and light trucks, heavy trucks, buses,
engines, and motorcycles.
Non-Road Mobile Source: Pollutants emitted by combustion
engines not associated with highway vehicles, such as farm and
construction equipment, gasoline-powered lawn and garden
equipment, power boats and outboard motors, and aircraft.
Biogenic Sources: Emissions produced by living organisms,
such as a forest that releases hydrocarbons.
Sources: Texas Commission on Environmental Quality, 2008;
U.S. EPA, 2008.
are available, states can use them in their assessment
of clean energy policies rather than develop a new
baseline emissions inventory. Sources of completed
state and local inventories that states and localities can
adopt for use in their analyses include:
EPA State GHG Inventories: EPA maintains a Web
site on state GHG inventories, which includes a
table of state CO2 emissions from fossil fuel con-
sumption by sector, http://epa.gov/climatechange/
emissions/'state_energyco2inv.html
Links to maps and summaries of existing state-com-
piled greenhouse gas inventories are also available on
this Web site.7 http://epa.gov/climatechange/emissions/
state_ghginventories.html
Local Government Inventories. Many local gov-
ernments have compiled GHG and/or criteria
air pollutant inventories through the auspices of
ICLEIs Cities for Climate Protection or the U.S.
7 State CO2 estimates are based on state energy data from the Energy Infor-
mation Administration (EIA), which maintains a database of state energy-
related data including fuel consumption by sector, electricity consumption, and
forecasts of the electric generation sector (U.S. DOE, 2008b).
EMISSIONS FACTOR APPROACH
An emissions factor quantifies the amount of a pollutant
released to the atmosphere from a "unit" of an activity or
source (e.g., Ibs CO2 per therm CH4 burned). The emissions
estimates are calculated by multiplying the emissions factor
(e.g., pounds of NOX per kWh produced) by the activity level
(e.g., kWh produced). Emissions factors can be calculated
based on the chemical composition of the fuels burned or
determined by emissions monitors.
Emissions factors for CO2, NOx, SO2, and other pollutants are
available from:
EPA's Emissions Factors and Policy Applications Center
http://www.epa.goV/ttn/chief/efpac/.html
EPA's Emissions & Generation Resource Integrated
Database(eGRID)
http://www.epa.gov/egrid
EPA's U.S. Greenhouse Gas Inventory Reports
http://www.epa.gov/climatechange/emissions/
usinventoryreport.html
Intergovernmental Panel on Climate Change Emissions
Factor Database (EFDB)
http://www.ipcc-nggip.iges.or.jp/EFDB/main.php
Conference of Mayor s Climate Protection Agree-
ment. These inventories have typically been devel-
oped using the CACPS Tool described below. Many
of these local inventories can be found online.
National Emissions Inventory (NEI). States can use
the NEI to help establish an inventory of criteria
and hazardous pollutants. EPA prepares a national
database of air emissions information with input
from numerous state and local air agencies, tribes,
and industry. The database contains information
on stationary and mobile sources that emit criteria
air pollutants and their precursors, as well as
hazardous air pollutants (HAPs). The database also
includes estimates of annual emissions, by source,
of air pollutants in each area of the country. The
NEI includes emission estimates for all 50 states,
the District of Columbia, Puerto Rico, and the
Virgin Islands, and is updated every three years.
http://www.epa.gov/ttn/chief/eiinformation.html
If existing baseline inventories are not available, states
can develop their own using methods and tools de-
scribed below.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 100
-------
Approaches to Developing a Baseline
Emissions Inventory
There are two basic approaches for developing state
emissions inventories for criteria air pollutants and/
or GHGs: top-down and bottom-up approaches. Both
inventory approaches require energy use estimates and
emissions factors to convert estimates of energy use into
estimates of emissions, as described in the text box Emis-
sions Factor Approach. Top-down and bottom-up ap-
proaches vary in their level of data and aggregation and
can serve different purposes. While the inventory devel-
opment process can be time- and resource-intensive, it
does not necessarily entail complex modeling methods.
Table 4.2.2, Comparison ofTop-Down and Bottom-Up
Approaches for Developing a Baseline Air and/or GHG
Emissions Inventory and Projection, compares the key
aspects of top-down and bottom-up approaches. The
following section presents information about each
approach for developing an emissions inventory,
including their strengths and weaknesses, appropriate
applications, relevant data sources and resources, and
the tools available to states. Methods and approaches
for projecting inventories out into the future are also
described. For further information on described tools,
see the Information Resource Description table at the
end of this chapter.
Top-Down Inventory Development
A top-down inventory contains aggregated activity
data across the state or community, and is used to gen-
erate state-wide estimates of emissions of GHGs or cri-
teria air pollutants. For example, a top-down inventory
might report emission estimates for categories such as
an industry within a state; it would not contain data on
emissions from specific facilities or buildings.
TABLE 4.2.2 COMPARISON OF TOP-DOWN AND BOTTOM-UP APPROACHES FOR DEVELOPING A
BASELINE AIR AND/OR GHG EMISSIONS INVENTORY AND PROJECTION
When to Use this
Tools Protocols Advantages Disadvantages Method
Top-Down Inventory
EPA's State Inventory
Intergovernmental panel
Tool for GHGs. on Climate Change.
National Association EPA's Emissions
Can capture all
Does not provide
emissions in a state. in-depth sectoral
Reliable data are
of Clean Air Inventory Improvement available for most
Agencies (NACAA) Program.
and International
Council for Local
Environmental
Initiatives (ICLEI)
Clean Air and Climate
Protection Software
(community- or state-
major sources.
emission detail.
Use of state average
State-wide estimates
of emissions.
State-wide GHG
inventories.
factors may lead to Area SOUrce emission
some uncertainty or estimates for criteria
error in estimates. air pollutants.
Lacks spatial
resolution needed for
air quality modeling.
wide inventory).
Bottom-up Inventory
. NACAA and ICLEI's EPA Climate Leaders
Can provide more Requires highly
Clean Air and GHG Inventory Protocol. detailed or nuanced disaggregated
Climate Protection The World Resources profile of emissions. data which may be
Software (government institute (WRI) and World
operations inventory). Business Council on Sus-
Allows analysis of
indirect emissions
Sector-specific GHG
inventories.
Stationary source
difficult to obtain. emission estimates
May not capture all for criteria air
Emission Reporting tainable Development sources (purchased emissions in a state. pollutants.
Data (e.g.. Acid (WBCSD) GHG Protocol. electricity, etc).
Rain Program Data, California Registry
or facility specific
emission reports).
Protocols.
The Climate Registry.
When data required
for top-down
inventory are not
available.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 101
-------
GHG REGISTRIES
GHG registries are systems for quantifying and reporting
GHG emissions and/or activities to reduce emissions that are
developed by collaborations of organizations, such as states or
firms. By establishing consistent emission reporting protocols, a
registry provides a common framework for entities to complete
a GHG inventory of their own emissions or emissions reductions
and a credible repository for the data over time. Such a
collection of entity-level emissions data can help inform a state's
understanding of emission sources and activities being taken to
reduce emissions. A registry does not serve the same function
as an inventory since it does not provide a comprehensive or
complete set of data on all emissions sources.
Examples of registry efforts are:
The Climate Registry is a collaboration among states,
provinces, and tribes to develop a common greenhouse
gas emissions registry system across multiple
governments. Corporations with operations in multiple
states will be able to report emissions using a consistent
reporting protocol and management system. http://www.
theclima teregis try. org/.
The California Climate Action Registry (CCAR) was
established by California statute as a registry for GHG
inventories for corporate reporting within the state. CCAR
has developed a general protocol and additional industry-
specific protocols that give guidance on how to inventory
GHG emissions for participation in the Registry, http://
www. clima teregis try. org/
The Voluntary Reporting of CMC Program is a mechanism
by which corporations, government agencies, individuals,
and organizations can report their GHG emissions,
emission reductions, and sequestration activities to the
federal Energy Information Agency. It was established
under Section 1605(b) of the Energy Policy Act of 1992.
http://www.eia.doe.gov/oiaf/1605/index.html
EPA's Mandatory GHG Reporting Rule, as requested by
Congress under the FY2008 Consolidated Appropriations
Act, became effective December 29, 2009. It requires
sources above certain threshold levels monitor and
report GHG emissions and applies to fossil fuel suppliers
and industrial gas suppliers, direct GHG emitters and
manufacturers of heavy-duty and off-road vehicles and
engines., http://www.epa.gov/climatechange/emissions/
ghgrulemaking.html
Because the spatial characteristics of criteria air pollut-
ants are important, an ideal inventory would include
very detailed, source-specific data that can be used in
air quality modeling. However, some sources, such as
area sources (e.g., residential fuel use and industrial use
of paints, solvents, and consumer products), cannot
be easily attributed to individual sectors or sources
and lend themselves more appropriately to a top-down
approach.8 See the text box Sources of Air Pollution
Emissions above for a summary of the different sources.
While there may be circumstances where a state desires
significant detail about the sources of its GHG emis-
sions, GHG inventories do not require the same level
of detailed spatial resolution since, as described above,
a ton of GHGs in one part of the state affects global
climate change in the same way as a ton of the same
GHG in another part of the state. For GHG emission
inventories, the top-down approach is most appropri-
ate when developing state-wide estimates of emissions
and developing emission reduction targets.
Protocols
It is important to develop an inventory that adheres
to a comprehensive and detailed set of methodologies
for estimating emissions. For GHG emissions, these
methodologies are usually derived from standards
established by the Intergovernmental Panel on Climate
Change (IPCC, 2008). Specific methods, tools, and
protocols for developing top-down baseline GHG
emissions inventories, forecasting future emissions,
and tracking changes are available at both the state and
local levels. For criteria air pollutants, these method-
ologies are usually derived from standards established
by EPA's Emissions Inventory Improvement Program
(EIIP), which offers guidance for developing invento-
ries of criteria and hazardous air pollutants and green-
house gas emissions (EPA, 2007). The protocols vary
depending on the type of inventory data a state collects.
Data Needs
To complete a top-down state-wide energy-related
emissions inventory, a state needs a variety of data,
such as state-wide electricity generation; energy
consumption by sector; and coal, oil, and natural gas
production and distribution.9 Many of these data are
available from national sources, such as the Energy
Information Agency (EIA) State Energy Data System
(U.S. DOE, 2008a). Data on economic activity and hu-
man population levels may be needed to supplement
data sources. These data are also available from nation-
al sources such as the Bureau of Economic Analysis'
8 Mobile sources are included as a separate category from area sources in
typical air pollution inventories.
9 To expand the inventory beyond energy, states would need data on sources
such as agricultural crop production, animal populations, and fertilizer use;
waste generation and disposal methods; industrial activity levels; forestry and
land use; and wastewater treatment methods.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 102
-------
Regional Accounts and the Census Bureau Population
Estimates. Some tools, such as the State Inventory Tool,
described below, provide default values states can use.
Additional sources are described later in this section.
Tools
Tools to help state and local governments develop GHG
and criteria air pollutant emission inventories include:
EPA's State Inventory Tool. States can use EPAs State
Inventory Tool to develop top-down GHG inven-
tories. This interactive spreadsheet software tool
is based on IPCC guidelines. States can enter their
own data or use pre-loaded state-specific emissions
factors and activity levels from federally managed
databases, such as EPAs eGRID (http://www.epa.
gov/egrid) and DOE's EIA. The State Inventory Tool
can calculate GHG emissions from energy con-
sumption as well as from industrial processes, agri-
culture, forestry, and waste management. This tool
is generally used to develop state-wide inventories
that can be tracked over time, to determine sectors
a state might target for reductions and to measure
long-term progress against state-wide or communi-
ty-wide goals over time. The State Inventory Tool is
designed to generate inventories for each year in a
time series (currently 1990-2006). http://www.epa.
gov/climatechange/emissions/state_guidance.html
Clean Air and Climate Protection Software Tool. Lo-
cal governments can use the Clean Air and Climate
Protection Software (CACPS) tool to develop a
top-down inventory of both criteria air pollutants
and GHGs associated with electricity, fuel use, and
waste disposal. CACPS is a Windows-based soft-
ware tool and database developed by the National
Association of Clean Air Agencies (NACAA)10 and
the International Council for Local Environmental
Initiatives (ICLEI), with EPA funding. The 2005
version of the tool is provided free to state and
local governments. More recent versions can be
purchased from ICLEI.
While available to state as well as local governments,
the CACPS tool is most appropriate for developing
locality-wide or government operations GHG in-
ventories based on IPCC guidelines with the inclu-
sion of criteria air pollutants. The CACPS tool:
10 Formerly the State and Territorial Air Pollution Program Administrators
and Association of Local Air Pollution Control Officials (STAPPA/ALAPCO).
is based on end-use energy consumption and
excludes agriculture, forestry, industrial, and
energy production;
requires users to complete each inventory year
separately; and
allows for analysis of indirect emissions (e.g.,
electricity imported from another state, waste
sent to out-of state landfills).
It is important to note, however, that CACPS does
not include location-specific criteria air pollutant
inventories and so it is difficult to interpret air
quality impacts, http://www.cacpsoftware.org/
Bottom-up Inventory Development
While top-down inventories are developed using high-
level, aggregated energy and economic information,
bottom-up inventories are built from source, equip-
ment population, and activity data. Bottom-up inven-
tory development involves collecting information on
source number and type from individual entities (e.g.,
businesses, local governments) within the state. This
approach can supplement state-wide GHG and other
air pollutant emission inventories by providing ad-
ditional, more detailed information. Data collected in
this manner may provide a more accurate estimate of
emissions within particular sectors (e.g., state-owned
government buildings). A more detailed and time con-
suming method than the top-down approach, bottom-
up inventory development provides comprehensive
estimates of precursor emissions and details regarding
spatial and temporal attributes that are required for air
quality modeling applications.
For criteria air pollutant inventories, bottom-up
inventories are most appropriate for developing more
accurate estimates for on-road, non-road, and station-
ary source emissions that can easily be attributed to
individual sectors or sources (e.g., major industrial
and commercial emission sources, such as electricity
generators, manufacturing processes and chemical
processes). For GHG emission inventories, the bottom-
up approach is most appropriate when developing
sector-specific inventories, when the data required for
a top-down inventory are not available, or to provide a
better match when evaluating multi-pollutant controls.
Protocols
As with the top-down inventory, it is important to
develop a bottom-up inventory that adheres to a
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 103
-------
comprehensive and detailed set of methodologies for
estimating emissions. For GHG emissions, there are
several protocols that states can use, including:
EPA Climate Leaders GHG Inventory Protocol. The
Climate Leaders Protocol includes overall guidance
to corporations in the Climate Leaders Partnership
on issues such as denning inventory boundaries,
identifying GHG emission sources, denning and
adjusting a base year, reporting requirements, and
goal-setting. http://www. epa.gov/climateleaders
* The GHG Protocol. The GHG Protocol is a joint
effort of the World Resources Institute (WRI) and
the World Business Council on Sustainable Devel-
opment (WBCSD). The protocol was designed for
corporate inventories, but can be adapted for use
by state governments quantifying emissions from
their own operations. The protocol provides step-
by-step guidance on calculating GHG emissions
from specific sources (e.g., stationary and mobile
combustion, process emissions) and industry sec-
tors (e.g., cement, pulp and paper aluminum, iron
and steel, and office-based organizations), http://
www.ghgprotocol.org/
Local Government Operations Protocol for the
Quantification and Reporting of Greenhouse Gas
Emissions Inventories, released in September 2008.
The Local Government Operations Protocol was
created to help local governments develop consis-
tent and credible emission inventories based on
internationally accepted methods. Developed in
partnership by the California Air Resources Board,
California Climate Action Registry, ICLEI - Local
Governments for Sustainability, and The Climate
Registry, it involved a multi-stakeholder technical
collaboration that included national, state, and lo-
cal emissions experts, http://www.icleiusa.org
For criteria air pollutants, methodologies are usu-
ally derived from standards established by EPAs
EIIP program, which offers guidance for develop-
ing inventories of criteria and hazardous air pollut-
ants, http:/7www. epa.gov/ttn/chief/eiip/techreport/
Data Needs
Bottom-up inventories are data-intensive. Often data
are not as readily available from national databases as
for top-down inventories and thus may require a sig-
nificant level of effort and time to collect. To conduct
a bottom-up GHG inventory of the utility sector, for
example, a state would collect data on the fossil fuel
consumption of every electricity production site in the
state and convert it to GHG quantities based on the
carbon content of the specific fuels that were used. Al-
ternatively, for sources for which data exist, a state can
gather and analyze continuous emissions monitoring
(GEM) data for electric utilities.
If a state is interested in developing an inventory of its
operations-related emissions, it would collect and com-
pile data on its energy and electricity use, process emis-
sions, waste generated, and other emissions-generating
activities. These data are often obtained from utility
bills, fleet records, and similar records.
Bottom-up criteria air pollutant inventories typically
use data gathered through surveys and reports from
emission sources, source permits, stack test data, and
GEM data. As described above, while obtaining data
can be difficult, the bottom-up approach can yield a
more detailed or nuanced profile of emissions for a par-
ticular sector than a top-down approach. More infor-
mation about existing data sources is provided below.
Tools
States can use a variety of tools to help develop
bottom-up GHG and criteria air pollutant inventories.
For GHG inventories:
Portfolio Manager is a free, interactive ENERGY
STAR energy management tool that enables users
to track and assess energy and water consump-
tion for a single building or across a portfolio of
buildings. A new feature of Portfolio Manager
lets users see how their buildings' CO2 emissions
compare with other buildings in the same region
and across the country, and measure their prog-
ress in reducing emissions. The tool can be used
to identify buildings with the most potential for
energy efficiency improvements, http://www.en-
ergystar.gov/index. cfm ?c= evaluate_performance.
bus_portfoliomanager_carbon
For criteria air pollutant inventories:
Point Sources: Most criteria pollutant inventories
for point sources are developed from permits and
other facility data rather than from a series of tools.
One tool that may complement this approach is the
Landfill Gas Emissions Model (LandGEM), a free,
automated estimation tool with a Microsoft Excel
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 104
-------
interface that can be used to estimate emission rates
for total landfill gas, methane, CO2, nonmethane
organic compounds, and individual air pollutants
from municipal solid waste landfills. http://www.
epa.gov/ttn/catc/dirl/landgem-v302-guide.pdf
Mobile Sources: Inventories for on-road and non-
road mobile sources can be aided by tools such as:
MOBILE6, a computer program that estimates
emission rates for mobile pollutants such as
hydrocarbon (HC), carbon monoxide (CO),
oxides of nitrogen (NOX), exhaust particulate
matter (which consists of several components),
tire wear parti culate matter, brake wear par-
ticulate matter, sulfur dioxide (SO2), ammonia
(NH3), six hazardous air pollutants (HAPs),
and carbon dioxide (CO2). MOBILE6 focuses
on gasoline-fueled and diesel highway motor
vehicles, and for certain specialized vehicles
such as natural-gas-fueled or electric vehicles
that may replace them. MOBILE6 uses county
or link-level VMT, speed, registration, and
roadway classification data to estimate emis-
sions from motor vehicles, http://www.epa.
gov/OMS/m6.htm
NON ROAD 2005 calculates past, present, and
future emission inventories (i.e., tons of pol-
lutant) for all nonroad vehicle and equipment
categories (e.g., recreational vehicles, agricul-
tural equipment, industrial equipment) except
commercial marine, locomotives, and aircraft.
The fuel types included in the model are gaso-
line, diesel, compressed natural gas, and lique-
fied petroleum. The model estimates exhaust
and evaporative HC, CO, NOx, particulate
matter, SO2, and CO2 emissions. The user can
select a specific geographic area (i.e., national,
state, or county) and time period (i.e., annual,
monthly, seasonal, or daily) for analysis. The
NONROAD tool includes estimates of equip-
ment population and activity and appropriate
emissions factors to estimate emissions from
these types of sources, http://www.epa.gov/
oms/nonrdmdl.htm
Motor Vehicle Emission Simulator (MOVES) is
a replacement for MOBILE6 and NONROAD
that EPA is currently developing. This new
emission modeling system will estimate emis-
sions for on-road and nonroad mobile sources,
cover a broad range of pollutants, and allow
multiple scale analysisfrom fine-scale analy-
sis to national inventory estimation. When
fully implemented, MOVES will serve as the
replacement for MOBILE6 and NONROAD
for all official analyses associated with regula-
tory development, compliance with statutory
requirements, and national/regional inventory
projections, http://www.epa.gov/otaq/models/
moves/index.htm
Data Sources and Additional Resources
for Top-Down and Bottom-Up Inventories
Many sources of data exist that states can use as they
compile top-down or bottom-up inventories. Some of
these data sources focus specifically on criteria air pol-
lutants, some focus on GHGs, and some include both.
Other sources provide already-compiled emissions
estimates. These resources are listed in Table 4.2.3 and
described below.
Emissions & Generation Resource Integrated Data-
base (eGRID). This free, publicly available software
from EPA has data on annual SO2, NOx, CO2, and
Hg emissions for most power plants in the United
States. eGRID also provides annual average non-
baseload emission rates, which may better charac-
terize the emissions of load-following resources.11
By accessing eGRID, states can find detailed emis-
sions profiles for every power plant and electric
generating company in the United States, http://
www. epa.gov/egrid
Emissions Collection and Monitoring Plan Sys-
tem (ECMPS). EPA collects data in five-minute
intervals from Continuous Emissions Monitors
(GEMS) at all large power plants in the country.
The ECMPS is a new system of reporting emissions
data, monitoring plans, and certification data, and
replaces the Emission Tracking System (ETS) that
previously served as a repository of SO2, NOx,
and CO2 emissions data from the utility industry.
http://www.epa.gov/airmarkets/business/
WRI Climate Analysis Indicators Tool: The Climate
Analysis Indicators Tool (CAIT-US.) provides a
free, comprehensive, and comparable database of
GHGs and other climate-relevant indicators for
U.S. states, http://cait.wri.org/
11 "Load-following" refers to the order in which different types of generating
equipment are used to meet changing electricity demand.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 105
-------
TABLE 4.2.3 SOURCES OF AIR POLLUTANTS AND GHG EMISSIONS DATA, INVENTORIES
| Type of Air Pollutant or GHG Emissions Approach
^^^K limmMH
National Emissions Inventory (NEI)
eGRID
Emissions Collection and Monitoring Plan System (ECMPS)
World Resources Institute Climate Analysis Indicators Tool
EPA State GHG Inventories
X
X
X
X
X
X
Local GHG Inventories
X
X
X
X
X
X
X
X
X
Top- Bottom-
" Down Up
X
X
X
X
X
X
X
X
X
X
State Agencies and Universities: Many state agencies
and universities collect emissions and/or energy
data within their state, which can be compiled into
an inventory.
Forecasting Future Emissions
To conduct a prospective analysis of potential emis-
sion reductions from a future policy, it is necessary to
develop forecasts of both the new policy case and the
"business as usual" (BAU) case that does not include the
new policy12 Emission projections provide a basis for:
Developing control strategies for State Implemen-
tation Plans (SIPs) or mitigation measures for
Climate Change Action Plans;
Conducting air quality attainment analyses; and
Tracking progress toward meeting air quality stan-
dards or GHG reduction goals.
When developing emission projections, an attempt is
made to account for as many of the important variables
that affect future year emissions as possible. States can
project future emissions based on historic trends and
expectations about numerous factors, including projec-
tions of population growth and migration, economic
growth and transformation, fuel availability and prices,
technological progress, changing land-use patterns,
12 When conducting a prospective analysis of clean energy policies that have
already been implemented, a forecast of emissions is not necessary although
it could facilitate projecting the future benefits of existing programs. For a
retrospective analysis, the impacts of the existing clean energy program could
be backed out of the forecast and reintroduced to estimate the impacts.
and climate change.13 The degree to which any of these
specific drivers is important is a function of the projec-
tion horizon. For example, climate change impacts may
be negligible for a five- to ten-year projection.
Several guidance documents and tools are available to
help states understand methodologies and data sources
for factors relevant to projections, including:
EPA EIIP Technical Report Series, Volume X: Emis-
sions Projections. This document provides informa-
tion and procedures to state and local agencies
for projecting future air pollution emissions for
the point, area, and onroad and nonroad mobile
sectors. It describes data sources and tools states
might use for their projections, http://www.epa.
gov/ttn/chief/eiip/techreport/volumel 0/xOl .pdf
EPA State GHG Projection Tool. States can use this
EPA spreadsheet tool to create forecasts of BAU
GHG emissions through 2020. Future emissions
are projected using a combination of linear extrap-
olation of the results from the State Inventory Tool,
described above, combined with economic, energy,
population, and technology forecasts. The tool can
be customized, allowing states to enter their own
assumptions about future growth and consumption
patterns, http://www.epa.gov/climatechange/wycd/
stateandlocalgov/analyticaltools.html
13 Some of these factors are closely related, and will rely on specific compo-
nents of these trends that may include electricity imports and exports, power
plant construction or retirement, domestic vs. imported agricultural produc-
tion, waste production, number of road vehicles, tons of freight transported,
vehicle miles traveled, and environmental regulations.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 106
-------
TABLE 4.2.4 COMPARISON OF BASIC AND SOPHISTICATED APPROACHES FOR QUANTIFYING AIR
POLLUTANT AND GHG EMISSION EFFECTS OF CLEAN ENERGY INITIATIVES
eCalc
OTC Workbook3
CACPS
Sophisticated Approaches
ENERGY 2020
NEMS
. IPM
MARKAL
PROSYM
GEMAPS
PROMOD
Advantages
lisadvantage:
/hen to Use this Method
Transparent.
Modest level of time,
technical expertise, and labor
required.
Inexpensive.
May be imprecise.
May be inflexible.
May have embedded
assumptions that have large
impacts on outputs.
Preliminary studies for short-
term resource planning.
Designing new programs and
evaluating existing ones.
Regulatory compliance and
energy plans.
More rigorous than basic
modeling methods.
May be perceived as more
credible than basic modeling
methods.
Allows for sensitivity analysis.
May explicitly account for
and quantify leakage.
Less transparent than
spreadsheet methods.
Labor- and time- intensive.
i Often high software licensing
costs.
Requires assumptions that
have large impact on outputs.
May require significant
technical experience.
State Implementation Plans.
Late-stage resource
planning.
Rate cases.
Project financing.
Regulatory compliance and
energy plans.
' The OTC workbook is a spreadsheet tool that was developed from specific results of the PROSYM model.
The Clean Air and Climate Protection Software Tool.
As described above, states or localities can use this
tool to project an emissions baseline of GHGs and
criteria air pollutants into the future, and measure
the effects of different policies upon the forecast.
http://www.icleiusa.org/cacp
States can also project future emissions based on their
energy baseline projections. More information about
forecasting energy baselines is available in Chapter 2,
Assessing the Potential Energy Impacts of Clean Energy
Initiatives.
4.2.2 STEP 2: QUANTIFY AIR AND GHG
EMISSION REDUCTIONS FROM CLEAN
ENERGY MEASURES
Once states have developed their baseline emission esti-
mate or business as usual forecast, they can estimate the
emissions that are avoided when implementing clean
energy measures. Although an emission reduction esti-
mation can be performed independently from a baseline
emissions forecast, aligning many of the assumptions in
the baseline case and the clean energy measures case is
a desirable exercise. Table 4.2.4 shows that states can use
either basic or sophisticated approaches to quantify air
emission reductions from clean energy measures.
Basic approaches typically include spreadsheet-based
analyses that use emissions factor relationships or
other assumptions to estimate reductions. Sophisti-
cated approaches are usually more complex and involve
dynamic electricity or energy system representations
that predict energy generation responses to policies
and calculate the effects on emissions. (For more spe-
cific information on these energy-related models, see
Chapters 2 and 3.)
Key Considerations for Selecting an Approach
for Quantifying Emission Reductions from
Clean Energy
As summarized in Table 4.2.4, there are advantages and
disadvantages to each approach for quantifying emis-
sion reductions. States can use this information as guid-
ance in determining the most appropriate approach for
their particular goals. It is important for states to:
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 107
-------
Consider the cost of each potential approach and/
or tool and the resources required;
Determine whether the tools or methods can be
used to estimate the pollutants and emissions of
interest;14 and
Decide between a complex, detailed approach and
a simple, transparent screening-level approach
based on their pros and cons and relative impor-
tance of each.
Basic and sophisticated approaches, including associ-
ated uncertainties and limitations, are described in
greater detail below.
Basic Approaches to Quantifying Emission
Reductions
Basic, screening-level, approaches involve: 1) establish-
ing the operating characteristics of the clean energy
resource, also known as its load profile; 2) identifying
the marginal generation unit and developing avoided
emissions factors; and 3) calculating the total emissions
reductions by multiplying the avoided emissions factor
by the avoided electricity generation (i.e., as calculated
in Chapter 2, Assessing the Potential Energy Impacts of
Clean Energy Initiatives). These procedures are illus-
trated in the flowchart in Figure 4.2.1 and described in
greater detail below.
Step 2a: Establish Clean Energy Operating
Characteristics (Load Profile)
As previously discussed in Chapter 2, Assessing the
Potential Energy Impacts of Clean Energy Initiatives, the
first step when applying a basic modeling approach is
to determine the specific ways that the clean energy
initiative will affect either demand for electricity or
available supply. This involves considering the follow-
ing issues related to the operating characteristics, or
load profile, of the clean energy measures:
How much energy will the clean energy mea-
sure generate or save? (See Chapter 2 for more
information)
When and where will the electricity generation
offset occur (e.g., season of year, time of day)? In
the case of energy efficiency measure, load impact
profiles describe the hourly changes in end use
14 The Model Energy Efficiency Program Impact Evaluation Guide, which
was developed as part of the National Action Plan for Energy Efficiency
(NAPEE), provides further guidance on how to quantify emissions reductions
(NAPEE, 2007).
FIGURE 4.2.1 BASIC APPROACHES FOR
QUANTIFYING AIR AND GREENHOUSE GAS
REDUCTIONS FROM CLEAN ENERGY
STEP 2A
Establish Clean Energy
Operating Characteristics
(Load Profile)
STEP 2B
i
Identify Marginal Generation
Unit and Develop Emissions
Characteristics
STEP 2C
Calculate Total Emissions
Reductions
OPTION 2B.1
Regional or system
average factors
OPTION 2B.2
Unit type factors
OPTION 2B.3
Load duration curve
derived factors
demand resulting from the program or measure.
In the case of energy resources, the generation
profiles (for wind or PV, for example) are required.
(See Chapter 3)
What, if any, are the emissions characteristics of the
clean energy resource (e.g., emissions characteris-
tics of using renewable fuels such as digester gas)?
Step 2b: Identify the Marginal Generation Unit
and Develop Emissions Characteristics
Next, identify the marginal generation source and its
associated emissions characteristics. The marginal gen-
erating source, as described earlier, is the last generating
unit to be dispatched in any hour, based on least-cost
dispatch (thus it is the most expensive on a variable cost
basis). The emissions characteristics of this unit can be
expressed as an emissions factor for each pollutant, and
are expressed in pounds per MWh. These factors rep-
resent the reduction in emissions per pound of energy
generation avoided due to energy efficiency or due to
clean energy resources supplied to the system.
There are several different approaches that can be used
to characterize the marginal generation source and its
associated emissions factor. As described in Chapter
3, these include (1) system average, (2) factors based
on unit type or other characteristic that correlates
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 108
-------
TABLE 4.2.5 COMPARISON OF METHODS TO IDENTIFY MARGINAL UNIT AND ASSOCIATED
EMISSIONS FACTOR
Method
Regional or system average
based on historical year
Advantages
Computationally simple.
Less labor and data required
than for unit type or dispatch
curve analysis.
Disadvantages
Insensitive to dispatch
process.
Neglects power transfers
between areas.
History may not be good
indicator of future.
When to Use this Method
Rough estimates of clean
energy benefits for displacing
emissions.
Based on unit type (capacity
factor rule)
Simpler and less labor
required than dispatch curve
analysis.
Considers generation
resource characteristics.
Somewhat insensitive to
dispatch process.
i Inaccurate for baseload clean
energy resources.
Preliminary planning and
evaluation of clean energy
resources, especially those
that operate during peak
times.
Derived from dispatch curve
analyses
More sensitive to dispatch
process than regional or
system average and unit type
methods.
Higher data requirements
than regional or system
average and unit type
methods.
Planning and regulatory
studies.
with likelihood of displacement (e.g., capacity factor),
and (3) factors derived from dispatch curve analyses.
Information about the advantages, disadvantages, and
when to use each method is summarized in Table 4.2.5,
Comparison of Methods to Identify Marginal Unit and
Associated Emissions Factor. Each method is described
in more detail below.
Regional or system average emissions factors. This
approach typically involves taking an average of the
annual emissions of all electricity generating units
in a region or system over the total energy output
of those units. Data on emission rates averaged by
utility, state, and region are available from EPA's
eGRID database. For example, using eGRID, states
can locate emissions factors by eGRID subregion,
state, or by specific boiler, generator, or plant.
While easy to apply, this method ignores the fact
that some units (such as baseload electricity gener-
ating units) are extremely unlikely to be displaced
by clean energy resources (see text box What
Energy Source is Displaced?). Baseload units and
other units with low variable operating costs (e.g.,
hydro and renewables) can be excluded from the
regional or system average to partially address this
shortcoming. Some approaches, therefore, take a
fossil-only average.
WHAT ENERGY SOURCE IS DISPLACED?
It is important to note that only a small number of
generating plants are affected by a clean energy measure.
Power systems are generally dispatched based on
economics, with the lowest-cost resource dispatched first
and the highest-cost resource dispatched last. The lowest-
cost units (known as baseload units) operate at all times
and are often fueled by coal. Higher-cost units such as
gas- and oil-fired units are brought online during peak use
times. These are the units that will be displaced by a clean
energy measure. This helps identify where the GHG and air
pollutant benefits are likely to occur (See Section 3.1, How
Clean Energy Can Achieve Electric System Benefits, and
Section 3.2, How States Can Estimate the Electric System
Benefits of Clean Energy, for a more detailed explanation
of how generation resources are dispatched).
Other methods for identifying the marginal unit
and its emissions factors attempt to recognize that
what is on the margin is a function of the time that
clean energy load impacts (or energy generation)
occurs. The most complete of these time-depen-
dent methods would analyze the impact of changes
in load for the 8,760 hours in a year using dispatch
models. Basic methods try to approximate this us-
ing proxies, including unit type and capacity factor,
as described further below.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 109
-------
FIGURE 4.2.2 CAPACITY FACTORS AND UNIT
DISPLACEMENT FOR BASELOAD AND LOAD-
FOLLOWING PLANTS
In general, baseload plants operate all of the time throughout
the year because their operating costs are low and because
they are typically not suitable for responding to the many
fluctuations in load that occur throughout the day. Thus, their
capacity factors are generally very high (e.g., greater than 0.8)
and they are unlikely to be affected by short-term fluctuations
in load. In contrast, load-following plants that can quickly
change output have much lower capacity factors (e.g., less
than 0.3) and are more likely to be displaced.
The capacity factor of a plant can be used as a proxy for how
likely the plant is to be displaced by a clean energy measure.
The following graph shows an example of a displacement
curve, or a rule for relating the likelihood that a unit's output
would be displaced to its capacity factor. Baseload plants
on the right side of the curve, such as nuclear units, are
assumed to be very unlikely to be displaced; peak load plants
on the left, such as combustion turbines, are much more likely
to be displaced.
Sample curve for relating displacement to capacity factor
'00=4
IE i)
601*
VY'.-r.
4'Jc'i
:: [I
- cj
Oc.;
~\
\
\
\
\..
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.3 0:3
Unit Capacity Factor
Source: Keith and Biewald, 2005.
Displaced unit and emissions factors identification
based on type of unit. As described above, system
or regional average emissions factors do not take
into account the fact that some electricity generat-
ing units are more likely to be displaced by clean
energy resources than others. (See Section 3.1, How
Clean Energy Can Achieve Electric System Benefits
and Section 3.2, How States Can Estimate the
Electric System Benefits of Clean Energy, for a more
detailed explanation of how generation resources
are dispatched.) The unit type approach for estimat-
ing emissions factors takes into account that some
classes of units are more likely to be displaced than
others by the operation of clean energy measures.
For example, assume coal, nuclear, and hydro plants
provide baseload power for an electricity grid.
Higher-cost units will operate in a cyclic manner,
increasing their output during peak daytime hours.
A more efficient new gas-fired unit may be counted
on to increase output during the day and decrease
output at night, while older, less efficient and more
expensive gas and oil units or combustion turbines
are only dispatched during the peak output periods.
This method can be made more representative by
disaggregating the unit types as much as possible
(e.g., by unit type, heat rate, and controls).
Estimating emissions factors based on unit type
involves the following steps.
1. Estimate the percentage of total hours each type
of unit (e.g., coal-fired steam, oil-fired steam, gas
combined-cycle, gas turbine, etc.) is likely to be
on the margin (the highest-cost unit dispatched
at any point in time is said to be "on the mar-
gin" and is known as the "marginal unit") and
thus to have its output displaced given the load
profile of the new clean energy resource. This
is discussed further in Chapter 3.
2. Determine the average emission rate for each
unit type (in pounds of emissions per MWh
output). This can be determined based on pub-
lic data sources such as EPAs eGRID database
or standard unit type emissions factors from
EPA AP-42, an available resource for estimated
emissions factors.15
3. Calculate an emissions-contribution rate for
each unit type by multiplying the unit type
average emissions (Ibs/MWh) by the fraction of
hours that the unit type is likely to be displaced.
Using average emissions to approximate displaced
emissions involves significant simplifications of
electric system operations. For example, the emis-
sion rates for each existing generating unit may vary
considerably. Similarly, plants of a certain type may
have different operating costs and load-following ca-
ls Note that AP-42 does not provide GHG emissions factors; for GHGs, use
fuel-specific emissions factors from EPAs Inventory of U.S. Greenhouse Gas
Emissions and Sinks. Also note that AP-42 factors are dependent on the air
pollution controls that have been installed, and this information would be
needed to accurately estimate emission rates. EPA AP-42 is available at http://
www.epa.gov/ttn/chief/ap42/index.html
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 110
-------
pabilities.16 For example, baseload units operate vir-
tually all the time, load-following units are routinely
turned off at night and used most days to meet
the higher daytime electricity demand, and peak-
ing units only operate during the highest demand
periods (such as hot summer afternoons). Due to
the operating characteristics of many types of clean
energy projects, the electricity produced or saved is
likely to displace electricity from load-following and
peaking units in the short term, rather than from
baseload units.17 Generalizations must also be made
about the type of generating unit that is on the mar-
gin, which may vary considerably across different
control areas and time periods.
A limitation of this approach is that it misses
important system-level dynamics. For example,
reducing emissions of a regulated pollutant may
result in shifts in other dispatch decisions in the
short and long term. This is particularly true if
those emission reductions have a market value (as
in cap and trade system). For example, if an energy
efficiency option allows for reduced output from a
high-emitting oil/gas steam unit during the shoul-
der period (i.e., that period when demand falls
below peak levels but above minimum, base load
levels), it may allow increased operation of a coal
plant (one not running at full utilization already)
at an increased capacity factor. This may reduce
system costs all while maintaining emissions at
capped levels. In other words, the clean energy op-
tion has allowed the operator to reduce emissions
compliance costs through dispatch changes. Over
the longer term these impacts may include changes
in retrofit or build decisions.
As an alternative to estimating the fraction of the
time each unit type is on the margin, some analy-
ses estimate the likelihood that a unit type could
be displaced using a displacement curve based on
capacity factors, shown in Figure 4.2.2, Capacity
Factors and Unit Displacement for Baseload and
Load-Following Plants. The capacity factor is the
ratio of how much electricity a plant produces to
how much it could produce, running at full capac-
ity, over a given time period. Historical data on, or
16 "Load-following" refers to those generating resources that are dispatched
in addition to baseload generating resources to meet increased electricity
demand, such as during daytime hours.
17 In the longer term, the electricity saved from EE or produced from CE proj-
ects not specific to time of day (e.g., CHP, geothermal, not solar) can displace
electricity from baseload resources.
estimates of, capacity factors for individual plants
are available from EPAs eGRID database.
Displacement rules do not capture some aspects
of electric system operations. For example, an
extended outage at a baseload unit (for scheduled
maintenance or unanticipated repairs) would
increase the use of load-following and peaking
units, affecting the change in net emissions from
the clean energy project. According to a displace-
ment rule, this plant would be more likely to be
displaced even though it would rarely if ever be on
the margin. Nevertheless, adding this level of detail
when estimating emissions factors will generally
produce a more credible and accurate estimate
of displaced emissions than relying simply on an
unweighted system average emissions rate.
Emissions Factors Derived from Dispatch Curve
Analyses Load curve analysis is a method for
determining tons of emissions avoided by a clean
energy resource for a period of time in the past.
In general, generating units are dispatched in a
predictable order that reflects the cost and opera-
tional characteristics of each unit. These plant data
can be assembled into a generation "stack," with
lowest marginal cost units on the bottom and high-
est on the top. A dispatch curve analysis matches
each load level with the corresponding marginal
supply (or type of marginal supply). Table 4.2.6,
Hypothetical Load for One-Week Period on Margin
and Emission Rate and Figure 4.2.3, A hypothetical
dispatch curve representing 168 hours by generation
unit, ranked by load level, provide a combined ex-
ample of a dispatch curve that represents 168 hours
(a one-week period) during which a hypothetical
clean energy resource would be operating.
Table 4.2.6 illustrates this process for a one-week
period. There are ten generating units in this
hypothetical power system, labeled 1 through 10.
Column [3] shows the number of hours that each
unit is on the margin, and column [4] shows the
unit's SO2 emission rate. The weighted average SO2
emission rate for these units is 5.59 Ib/MWh.
In many cases, dispatch curves are available from
the local power authorities and load balancing au-
thorities (e.g., a regional Independent System Op-
erator (ISO)). If this information is not available,
states can attempt to construct their own analysis.
Constructing a dispatch curve requires data on:
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 111
-------
TABLE 4.2.6 HYPOTHETICAL LOAD FOR
ONE-WEEK PERIOD: HOURS ON MARGIN
AND EMISSION RATE
1
2
3
4
5
6
7
8
9
10
Oil Combustion Turbine, Old
Gas Combustion Turbine
Oil Combustion Turbine, New
Gas Steam
Oil Steam
Gas Combined Cycle, Typical
Gas Combined Cycle, New
Coal, Typical
Coal, New
Nuclear
5
10
9
21
40
32
17
34
0
0
1.00
0.00
1.00
0.10
12.00
0.01
0.01
13.00
1.00
0.00
Weighted average, SO2 emissions (Ibs/MWh): 5.59
FIGURE 4.2.3 A HYPOTHETICAL LOAD
DURATION/DISPATCH CURVE REPRESENTING
168 HOURS (SHOWN IN HALF-DAY
INCREMENTS) BY GENERATION UNIT,
RANKED BY LOAD LEVEL
DOil Combustion Turbine, Old
Gas Combustion Turbine
Oil Combustion Turbine, New
DGas Steam
Oil aeam
Q Gas Combined Cycle, Typical
Gas Combined Cycle, New
nCoal, Typical
DCoal, New
Nuclear
1 13 25 37 49 61 73 85 97 109 121 133 145 157
Hour
Source: Developed by Synapse Energy, unpublished, 2007.
1. Historical utilization of all generating units in
the region of interest;
2. Operating characteristics, including costs and
emissions rates of the specific generating units,
for each season;
3. Energy transfers between the control areas of
the region and outside the region of interest
in order to address leakage issues (see text
box Clean Energy and Leakage earlier in this
chapter); and
4. Hourly regional electricity demand (or loads).
Data on operating cost, historical utilization, and
generator-specific emission rates can typically be
obtained from the EIA (http://www.eia.doe.gov/
cneaf/electricity/page/data.html), or the local load
balancing authority. When generator cost data are
not available, capacity factors (from the eGRID
database, for example) for traditional generating
units can be used to approximate the relative cost
of the unit (those with the highest capacity factors
are assumed to have the lowest cost). As an excep-
tion, variable power resources such as wind and
hydropower are assumed to have lower costs than
fossil fuel or nuclear units.
If unit-level cost data are available, calculating the
weighted average of each unit's emission rate, as
shown in Table 4.2.6, is preferable to aggregating
plants, especially when there is considerable varia-
tion in the emission rates within each unit type.
Operational data (or simplifying assumptions)
regarding energy transfers between the control
areas of the region and hourly regional loads can
be obtained from the ISO or other load balancing
authority within the state's region.
Load duration curve analysis is commonly used in
planning and regulatory studies. It has the advan-
tage of incorporating elements of how generation
is actually dispatched while retaining the simplicity
and transparency associated with basic model-
ing methods. However, this method can become
labor-intensive relative to other basic modeling
methods for estimating displaced emissions if data
for constructing the dispatch curve are not readily
available. Another disadvantage is that it is based
on the assumption that only one unit will be on the
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 112
-------
margin at any given time; this is not generally true
in most regions.
Summary of Emissions Factor Methods. In general,
for each of the three methodsregional or system
emissions factors, factors based on unit type, and
factors derived from load duration/dispatch curve
analysesthe more detailed the analysis, the
more accurate the results, but the more involved
it is to make the calculations. The accuracy of the
analysis can be improved by calculating separate
emissions factors for a number of different time
periods during which load and unit operations are
known to vary (e.g., peak and off-peak times in the
winter and summer months). Ideally, several years
of historical emissions and generation data would
be used in calculating the average emission rate.
For the latter two methods (i.e., emissions factors
based on unit types and derived from load dura-
tion/dispatch curve analyses), the number of hours
that the unit type is on the margin would also be
incorporated into the calculation.
Step 2c: Calculate Total Emissions Reductions
Total emission reductions are calculated by applying
the emissions factor developed during Step2b Identify
the Marginal Generation Unit and Develop Emissions
Characteristics to the clean energy resource's level of
activity, determined during Step 2a Establish Clean
Energy Operating Characteristics.
In the final analysis of net emission impacts, it is also
important to consider any GHG or criteria air pollution
emissions that a clean energy initiative might produce
during the production or generation of renewable fuels
(e.g., landfill gas, biomass generation). For example,
biomass generation releases about the same amount of
CO2 as burning fossil fuels. However, because biomass
is a fuel derived from organic matter, including, but
not limited to, wood and paper products, agricultural
waste, or methane (e.g., from landfills), these materials
are part of the natural carbon cycle and therefore do
not contribute to global warming. Thus, all biomass
CO2 emissions (including those from renewable meth-
ane) are assigned a value of zero because these organic
materials would otherwise release CO2 (or other green-
house gases) through decomposition.
Tools
Several tools that take a basic modeling approach to
estimating emissions reductions are available to states:
USING LOAD DISPATCH CURVE EMISSIONS FACTORS TO
ANALYZE THE EMISSIONS IMPACT OF WISCONSIN'S ENERGY
EFFICIENCY PROGRAMS
In 2004, the Wisconsin Department of Administration (DOA)
released an analysis of the air emission impacts of its Focus on
Energy efficiency program. The DOA's evaluation team used
a load dispatch curve analysis to estimate which generating
plants were "on the margin" during different time periods.
Using EPA's CEM data on historical plant operations and
emissions reported to EPA, emissions factors were developed
for the marginal generating units for different time periods
(e.g., peak and off-peak hours during winter and summer) for
NOx, SO2, and CO2). These factors were then used to analyze
the effects of different energy efficiency programs.
The study found that the marginal units' emission rates tend
to be higher during off-peak hours (particularly in winter) than
on-peak hours. The study suggests that energy efficiency
programs that cut energy consumption in Wisconsin when
system demands (and power supply costs) are low may
produce the greatest reductions in emissions. For more
information on Wisconsin's Focus on Energy program, see
Section 4.3.2, Wisconsin - Focus on Energy Program.
Source: Erickson et ai, 2004.
The Clean Air and Climate Protection Software
(CACPS) tool can be used to estimate emissions
reductions in addition to the functions already
mentioned above. ICLEI updated and re-released
this software in April 2009. Web site: http://www.
icleiusa.org/cacp
The OTC Workbook: The OTC Workbook is a free
tool developed for the Ozone Transport Commis-
sion to help local governments prioritize clean ener-
gy actions. The Workbook uses a detailed Microsoft
Excel spreadsheet format based on electric power
plant dispatch and on the energy savings of various
measures to determine the air quality benefits of
various actions taken in the OTC Region. This tool
is simple, quick, and appropriate for scenario analy-
sis. It can calculate predicted emission reductions
from energy efficiency, renewables, energy portfolio
standards (EPSs), and multi-pollutant proposals.
The tool contains two kinds of default emission rate:
system average (for assessing EPSs) and marginal
(for assessing displacement policies). Users can also
input their own data, http://www.otcair.org
Power Profiler: The Power Profiler is a Web-based
tool that allows users to evaluate the air pollution
and GHG impact of their electricity choices. The
tool is particularly useful with the advent of electric
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 113
-------
ELECTRIC ENERGY EFFICIENCY AND RENEWABLE ENERGY IN
NEW ENGLAND: THE OTC WORKBOOK
An analysis conducted by the Regulatory Assistance Project
(RAP) explains how energy efficiency and renewable energy
have led to many positive effects on the general economy,
the environment, and energy security in New England while
also quantifying these effects in several new ways. The report
assesses the air quality effects of efficiency and renewable
investments using the OTC Workbook tool. The analysis finds
that there is clear progress in reducing CO2 emissions from the
deployment of energy efficiency and renewable energy. The
projections by the OTC Workbook indicate that due to current
energy efficiency programs, 22.5 million tons of CO2 emissions
are avoided from 2000-2010.
Source: The Regulatory Assistance Project, http://www.raponline.org/
Pubs/RSWS-EEandREinNE.pdf
customer choice, which allows many electricity
customers to choose the source of their power.
http://www.epa.gov/cleanenergy/powerprojiler.htm
eCalc: eCalc is an online tool that identifies emis-
sion reductions from energy efficiency and renew-
able energy measures in the Electric Reliability
Council of Texas (ERGOT) region. The eCalc tool
incorporates both energy modeling (assessing the
energy saved by a given measure) and emissions
modeling (determining the emissions avoided by
those energy savings). The energy modeling capa-
bility is extremely robust and detailed, accounting
for a wide array of load types with weather normal-
ization. It also includes energy production profiles
for wind and solar power. Several states have
approached the Energy Systems Laboratory (ESL)
at Texas A&M University about developing other
versions of eCalc. While the underlying code can
be transferred, states will need to customize data
such as weather, geography, building standards,
emissions regulations, grid characteristics, and
other factors, http://ecalc.tamu.edu/
Note that many of these spreadsheet-based and other
tools rely on models to estimate the underlying emis-
sion rates. For example, the OTC Workbook relied on
runs of the PROSYM model to establish the emission
rates, and eCalc integrates several legacy models
depending on the users desired analysis type. These
tools thus have the same underlying concerns as those
raised earlier, such as being dependent on key driving
assumptions; to the extent that these tools and their
inputs are not regularly updated, these key assumptions
may no longer be applicable and relevant.
A RESOURCE FOR CALCULATED AVOIDED EMISSIONS:
THE MODEL ENERGY EFFICIENCY PROGRAM IMPACT
EVALUATION GUIDE
The Model Energy Efficiency Program Impact Evaluation Guide
provides guidance on model approaches for calculating energy,
demand, and emissions savings resulting from energy efficiency
programs. The Guide is provided to assist in the implementation
of the National Action Plan for Energy Efficiency's five key
policy recommendations and its Vision of achieving all cost-
effective energy efficiency by 2025. Chapter 6 of the report
presents several methods for calculating both direct onsite
avoided emissions and reductions from grid-connected electric
generating units. The chapter also discusses considerations for
selecting a calculation approach (NAPEE, 2007).
Limitations of Basic Approaches
Basic approaches for quantifying displaced emissions
are analytically simple and the data are readily avail-
able. However, they involve a less rigorous approach
than sophisticated modeling approaches; policy-
making and regulatory decisions typically require more
rigorous analysis. Basic approaches:
Are best suited for estimating potential emis-
sion reduction benefits for a relatively short time
frame (e.g., one to three years). Longer-term
analyses would require emissions factors that ac-
count for impacts on the addition and retirement
of energy sources over time and changes in market
conditions including environmental requirements.
Do not typically account for imported power, which
may be from generating units with very different
emissions characteristics than the units within
the region or system. These methods also do not
account for future changes in electricity import/
export patterns, which may change the marginal
energy sources during operation of the clean en-
ergy measure.
Do not account for the myriad factors that influ-
ence generating unit dispatch on a local scale. For
example, the emissions impacts of a clean energy
resource within a load pocket (an area that is served
by local generators when the existing electric sys-
tem is not able to provide service, typically due to
transmission constraints) would affect unit dispatch
very differently than measures in an unconstrained
region. Higher-cost units must be dispatched in a
load pocket because energy cannot be imported
from lower-cost units outside of the area.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 114
-------
For these reasons, use of basic approaches is often
limited to providing preliminary estimates of emis-
sion reductions and reporting approximate program
impacts data for annual project reports and program
evaluations that do not involve regulatory compli-
ance. Nevertheless, when using basic approaches it
is important to remember that the more detailed the
representation of the study area, the more precise and
reliable the emissions estimates.
Sophisticated Approaches to
Quantifying Emissions Benefits
Sophisticated modeling approaches, such as electric
dispatch and capacity planning models, can be used to
compare baseline energy and emissions forecasts with
scenarios based on implementation of clean energy
measures. Using sophisticated models to estimate
emissions that are displaced as a result of clean energy
measures generally results in more accurate estimation
of emission impacts than using the basic approaches,
but can be more resource-intensive.
Many of the models used to characterize or project
changes in electricity supply and demand also provide
estimates of the air pollution and GHG impacts associ-
ated with clean energy policies. Thus, by comparing
clean energy policy scenarios with the BAU case, they
facilitate quantification of emissions benefits. Two key
types of models used to estimate emissions are electric
dispatch models and capacity expansion (also referred
to as system planning or planning) models. An electric
dispatch model typically answers the question: how
will this clean energy measure affect the operations
of existing power plants? In other words, the model
quantifies the emission reductions that occur in the
short term. A capacity expansion model answers the
question: how will this clean energy measure affect
the composition of the fleet of plants in the future? A
capacity model typically takes a long-term view and
can estimate emission reductions from changes to the
electricity grid, rather than changes in how a set of
individual power plants is dispatched.
Some capacity expansion models include dispatch
modeling capability, although typically on a more
aggregate time scale than dedicated hourly dispatch
models. Models that address dispatch and capacity
expansion handle both the short and long term. These
models are summarized in Table 4.2.7, Comparison of
Sophisticated Modeling Approaches for Quantifying Air
and GHG Emission Effects of Clean Energy Initiatives,
and are described in more detail in Chapters 2 and 3).
4.2.3 STEP 3: QUANTIFY AIR QUALITY
IMPACTS
When criteria air pollutants are reduced through clean
energy measures, as determined under Step 2, the
ambient concentrations of both primary and second-
ary criteria air pollutants are also likely to be reduced.
Estimating air quality improvements associated with
emission changes is another step in a thorough analysis
of the benefits of clean energy initiatives.18
Modeling ambient air quality impacts can be a complex
task, however, requiring sophisticated air quality mod-
els and extensive data inputs (e.g., meteorology). Many
state and local government air program offices already
use rigorous air quality modeling approaches for their
State Implementation Plans, as required by the Clean
Air Act. These approaches, summarized below, can also
be used in evaluating clean energy benefits.
Approaches to Quantifying Air Quality Changes
Sophisticated computer models are often necessary to
prepare detailed estimates of the impact of emission
changes on ambient air pollution concentrations. There
are three broad types of relevant air quality models:
dispersion models, photochemical models, and receptor
models. All of these models require location-specific
information on emissions and source characteristics, al-
though they may represent photochemistry, geographic
resolution, and other factors to very different degrees.
Dispersion Models. Dispersion models rely on
emissions data, source and site characteristics (e.g.,
stack height, topography), and meteorological in-
puts to predict the dispersion of air emissions and
the impact on concentrations at selected down-
wind sites. Dispersion models do not include anal-
ysis of the chemical transformations that occur in
the atmosphere, and thus cannot assess the impacts
of emission changes on secondarily formed PM25
and ozone. These models can be used for directly
emitted particles (such as from diesel engines) and
air toxics. EPA currently recommends using either
18 "Concentrations" versus "emissions:" Ambientor surroundingair
concentration levels are the key measure of air quality and are based on the
monitored amount (e.g., in units of microgramsper cubic meter [ug/m3] or
parts per million [ppm]) ofapollutant in the air. Emission levels are based on
estimates and monitored measurements of the amount (e.g., in units of tons)
ofapollutant released to the air from various sources, such as vehicles and
factories. Some emissions travel far from their source to be deposited on dis-
tant land and water; others dissipate over time and distance. The health-based
standards (National Ambient Air Quality Standards) for criteria pollutants
are based on concentration levels. The pollutant concentration to which a
person is exposed is just one of the factors that determines if health effects
occurand their severity if they do occur (U.S. EPA, 2009).
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 115
-------
TABLE 4.2.7 COMPARISON OF SOPHISTICATED MODELING APPROACHES FOR QUANTIFYING AIR AND
GHG EMISSION EFFECTS OF CLEAN ENERGY INITIATIVES
Examples of models
Advantages
lisadvantage:
/hen to Use this Method
Electric Dispatch
PROSYM
GEMAPS
PROMOD
Provides very detailed
estimations about specific
plant and plant-type effects
within the electric sector.
i Provides highly detailed,
geographically specific,
hourly data.
Often lacks transparency.
May require technical
experience to apply.
Labor- and time- intensive.
Often high labor and
software licensing costs.
Requires establishment of
specific operational profile of
the clean energy resource.
Often used for evaluating
Specific projects in small
geographic areas.
Short-term planning (0-5
years), and Regulatory
proceedings.
Capacity Expansion or Planning
NEMS
IPM
ENERGY 2020
LEAP
Model selects optimal
changes to the resource
mix based on energy system
infrastructure over the long
term (10-30 years).
May capture the complex
interactions and feedbacks
that occur within the entire
energy system.
Provides estimates of
emission reductions from
changes to generation mix.
May provide plant specific
detail and perform dispatch
simultaneously (IPM).
Requires assumptions that
have large impact on outputs
(e.g., future fuel costs).
May require significant
technical experience to
apply.
Often lacks transparency
of spreadsheet due to
complexity.
Labor- and time- intensive.
Often high labor and
software licensing costs.
Long-term studies (5-25
years) over large geographical
areas such as:
State Implementation Plans,
Late-stage resource
planning.
Statewide energy plans, and
GHG mitigation Plans.
the AERMOD Modeling System or CALPUFF in
SIP revisions analysis for existing sources and for
New Source Review. Numerous other dispersion
models are available as alternatives or for use in a
screening analysis, http://www.epa.gov/scram001/
dispersionindex. htm
Photochemical Models. The second type of air
quality models are photochemical models. Pho-
tochemical models include many of the complex
physical and chemical processes that occur in
the atmosphere as gaseous emissions of different
chemicals react and form PM25 and ozone. These
models perform complex computer simulations,
and can be applied at a variety of scales from the
local to the global level. Photochemical models
include EPAs Community Modeling and Analysis
System (CMAQ) and the Comprehensive Air
Quality Model (CAMx). A range of photochemi-
cal-type air quality tools are also available for use
in assessing control strategies. One example is the
Modeled Attainment Test Software (MATS), a PC-
based software tool for SIP attainment demonstra-
tions recently developed by EPA. While MATS is
not an air quality model per se, it combines CMAQ
or CAMx results with monitor data to calculate
design values, http://www.epa.gov/scram001/
photochemicalindex.htm
Receptor Models. Receptor models can identify and
quantify the sources of air pollutants at a receptor
location. Unlike photochemical and dispersion air
quality models, receptor models do not use pol-
lutant emissions, meteorological data, and chemi-
cal transformation mechanisms to estimate the
contribution of sources to receptor concentrations.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 116
-------
TABLE 4.2.8 AIR QUALITY MODELS CURRENTLY RECOMMENDED BY EPA AND AVAILABLE AT EPA'S SCRAM
Model Acronym
Dispersion Models
Model Name
CALPUFF
AERMOD
EPA-approved version of the
California Puff Model
American Meteorological
Society/EPA Regulatory Model
Single source model with air chemistry for secondary formation. Can analyze
secondary formation of ozone and PM2.5.
Recommended single source model for direct dispersion modeling (no air
chemistry). Replaced Industrial Source Complex (ISC) family of models. Capable of
multiple and area source analysis.
Photochemical Models for both Ozone and PM2.5 ("One Atmosphere" models)
CAMx
CMAQ
Comprehensive Air Quality
Model with extensions
For ozone, particulate matter, inorganic and organic PM2.5/PM10, mercury and
other toxics.
Community Multi-Scale Air For ozone, fine particles, toxics, acid deposition, and visibility degradation.
Quality model
Receptor Models
CMB
UNMIX
PMF
Chemical Mass Balance
N/A
Positive Matrix Factorization
The EPA-CMB Version 8.2 uses source profiles and speciated ambient data to
quantify source contributions. Contributions are quantified from chemically
distinct source types rather than from individual emitters. Sources with similar
chemical and physical properties cannot be distinguished from each other by
CMB. Many of the source profiles, however, are outdated.
The EPA UNMIX model "unmixes" the concentrations of chemical species
measured in the ambient air to identify the contributing sources.
A form of factor analysis where the underlying co-variability of many variables
(e.g., sample to sample variation in PM species) is described by a smaller set of
factors (e.g., PM sources) to which the original variables are related. The structure
of PMF permits maximum use of available data and better treatment of missing
and below-detection-limit values.
Source: U.S. EPA, 2008c.
Instead, receptor models use the chemical and
physical characteristics of gases and particles
measured at the source and receptor to identify
the presence of, and to quantify source contribu-
tions to, receptor concentrations. These models are
therefore a natural complement to other air quality
models and are used as part of SIPs for identifying
sources contributing to air quality problems, http://
www.epa.gov/scram001/receptorindex.htm
Additional models are available and may be suitable
for clean energy benefits analysis. EPA's Support Cen-
ter for Regulatory Modeling (SCRAM) provides in-
formation about the latest versions of models, as well
as the status of current recommendations of models
for regulatory purposes. Examples of all three of these
types of models are available at SCRAM and are sum-
marized in Table 4.2.8, Air Quality Models Currently
Recommended by EPA and Available at EPA's SCRAM.
http://www.epa.gov/scram001/aqmindex.htm
Some states have developed air quality models tailored
to their specific region. These models are typically used
for air quality policy development purposes, or for air
quality forecasting as part of an air quality index alert
system. Such local or regional models are suitable for
conducting clean energy benefits analysis, and the
expertise and data needed by these models are often
available within a state. An example of such a tool
is the Assessment of Environmental Benefits (AEB)
modeling system, described in the text box, which is
currently configured for use by the southeastern states.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 117
-------
Recently, approaches have been developed that use
the output of photochemical and dispersion models
to create screening tools that can be used to quickly
evaluate expected responses to emissions changes. These
screening tools use information from a series of model
simulations in which precursor emissions are reduced
by specified amounts (e.g., 10 percent NOX, 20 percent
NOX, 10 percent VOC, 20 percent VOC, etc.) and the
responses by various pollutants (e.g., ozone) are assessed
for each simulation to create a pollutant "response sur-
face" for a given area. Once the series of simulations has
been completed for a particular region, the users can
use the tool to more readily identify the emission reduc-
tion options or scenarios that seem most promising
relative to their goals. For those scenarios identified by
the screening tool as potentially effective, the user can
then re-run the full model for the identified scenarios
to more accurately evaluate the spatial and temporal as-
pects of the expected response. Although these screen-
ing tools provide a quick way of evaluating the expected
response for a variety of scenarios, time and resources
are required to develop the initial response surface for
each pollutant and each given area of interest.
Examples of air quality screening tools include:
EPA Response Surface Modeling (RSM): RSM
is based on a new approach known as air qual-
ity metamodeling, which aggregates numerous
pre-specified individual air quality modeling
simulations into a multi-dimensional air quality
"response surface." RSM is a metamodel of an
air quality model developed using the Commu-
nity Multi-Scale Air Quality (CMAQ) Modeling
systemit is a reduced-form prediction model us-
ing statistical correlation structures to approximate
model functions through the design of complex
multi-dimension experiments. RSM has been suc-
cessfully tested and evaluated for PM2 5 and ozone,
respectively (U.S. EPA, 2006a).
EPA's Source-Receptor (S-R) matrix: The S-R matrix
is a reduced-form model based on a regional disper-
sion model, the Climatological Regional Dispersion
Model (CRDM), which provides the relationship
between emissions of PM25 or particle precursors
and county-level PM25 concentrations. The S-R ma-
trix is used to evaluate PM2 in the Co-Benefits Risk
Assessment (COBRA) screening model described
later in this chapter (U.S. EPA, 2006b).
ASSESSMENT OF ENVIRONMENTAL
BENEFITS MODELING SYSTEM
The Assessment of Environmental Benefits (AEB) modeling
system is a web-based tool designed for southeast states to
use in estimating the ozone and PM impacts of their energy
efficiency and renewable energy projects. This coupled energy-
air quality modeling system was developed for use in the SIP
development process.
AEB takes user-provided inputs of electricity impacts (efficiency
gains or net generation) of location-specific energy efficiency
and renewable energy projects and estimates the reduced
emissions and air quality improvements that will occur by the
avoided conventional electricity generation.
Source: Imhoff, 2006
Key Considerations When Selecting a Method
to Assess Air Quality Impacts
Air quality impact analyses enable clean energy policy
analysts to quantify current and future changes in the
concentration of ambient air pollutants that affect hu-
man health. When selecting an air quality model that
will comprehensively model either short- or long-term
changes in air quality, particularly in urban regions,
there are a number of modeling inputs and other fac-
tors to consider.
The Pollutants for Analysis. Deciding what pollut-
ants to model is a critical decision when selecting a
model. Directly emitted primary pollutantssuch
as CO, SO2, direct PM, and many air toxics-
require models capable of modeling dispersion and
transport (i.e., dispersion models). Secondarily
formed pollutants such as O3 and most PM25 are
formed by chemical reactions occurring in the
atmosphere with other pollutants. Secondary pol-
lutants are considerably more difficult to model,
requiring a model capable of handling the complex
chemical transformations (i.e., photochemical
models), as well as short and long-range transport.
Sources Affected. The number and types of sources
that result in emissions directly affect the selection
of an appropriate air quality model. A model that
is appropriate for modeling the impact of a single
generating facility with a tall smokestack would be
inappropriate for analysis of an initiative that would
affect electricity generation throughout the region.
Timeframe. Pollutants are further distinguished
by the exposure timeframe that is most relevant
to human health impactse.g., long-term average
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 118
-------
exposure vs. short-term daily or hourly exposure.
The impact assessment timeframe can be a key
factor in determining appropriate approaches
for modeling air quality impacts of clean energy
initiative-based emission reductions.
Data Availability and Resolution. Air quality
models require large amounts of input data de-
scribing a variety of characteristics of the energy-
environment system, including emission inventory
data, ambient air quality monitoring data, and
meteorological data.
Geographic Scope. Selecting the most appropriate
analytical tool to model air quality impacts also de-
pends upon the geographical scope of the analysis.
Modeling large geographical areas (e.g., a state or
a group of states) often requires a different model
than when modeling smaller areas (e.g., a city)
Meteorological and Topographical Complexities.
When structuring an air quality impact analysis, it
is also important to consider regional meteorologi-
cal and topographical conditions that may affect
the transport and chemical reaction of pollutants
within a region's atmosphere. Thus, it is important
to determine whether air quality models can ac-
count for these factors.
4.2.4 STEP 4: QUANTIFY HUMAN HEALTH
AND RELATED ECONOMIC EFFECTS OF AIR
QUALITY IMPACTS
A central question for many clean energy stakeholders
regards the negative human health effects that can be
avoided through clean energy-related emission reduc-
tions. Estimates of the numbers of avoidable health
impactsfrom reduced school absences and lost work
days to avoided premature deathshave become stan-
dard and powerful techniques to describe the benefits
of air-related programs. Quantifying the avoidable
health effects associated with clean energy initiatives is
an analytical step that typically builds on the estimates
of emission reductions and air quality changes. Health
research has established strong relationships between
air pollution and health effects ranging from fairly mild
effects such as respiratory symptoms and missing a day
of school or work, to more severe effects such as hos-
pital admissions, heart attacks, onset of chronic heart
and lung diseases, and premature death.
Presenting the benefits of clean air initiatives in such
tangible terms as reduced cases of health effects can be
a valuable analytical tool to help differentiate between
alternative program options, as well as a very effective
technique for communicating some of the most impor-
tant advantages of clean energy. This section describes
basic and sophisticated modeling approaches to estimate
the human health effects of air quality changes and the
monetary value of avoided health effects, a key compo-
nent of a comprehensive economic benefit-cost analysis.
Methods for Quantifying Human Health
Impacts
Estimating the health benefits of air quality improve-
ments can be achieved through basic or sophisticated
modeling methods. Basic modeling approaches use
results from existing studies, such as regional impact
analyses, to extrapolate a rough estimate of the health
impacts of a single new facility or clean energy initia-
tive. Sophisticated modeling approaches include
screening-level analytical models that can run quickly
on a desktop computer, and rigorous and complex
computer models that often run on powerful comput-
ers and involve a linked series of separate models. Basic
and sophisticated approaches are described below.
Basic Modeling Approach
A common basic modeling approach for quantifying
the human health effects of a clean energy initiative
involves determining the "health benefit value per ton
of emission" (also referred to as the benefit per ton,
or BPT) to estimate average monetized benefits of an
incremental change in pollutant or pollutant precursor.
This is a form of "benefits transfer" analysis, where
the results from an extensive analysis (e.g., a regional
control strategy for all coal-fired power plants within a
region) are used to approximate the effects of a smaller
project in the same region (e.g., a local clean power ini-
tiative). In effect, these metrics represent a composite
of the air quality modeling, health impacts estimation,
and valuation estimation steps used in more complex
models, such as the BenMAP model described below.
EPA has recently developed PM2 BPT estimates cat-
egorized by key PM25 precursors, source category, and
location of the county (Fann, 2008). Applying these
estimates simply involves multiplying the emission
reduction by the relevant BPT metric.
BPT measures are only first-order approximations
of the results that a rigorous analysis might estimate.
However, they can serve as pragmatic benefits analysis
tools and can be especially useful in assessing the
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 119
-------
benefits of small projects where it is impractical to con-
duct a complex analysis of each alternative.
The role of BPT benefit estimates varies: some states
develop these estimates as a useful "rule of thumb" used
during screening analysis when formal air quality mod-
eling analyses are impractical due to time and resource
constraints, while other states use the estimates as a
more formal part of the analysis of proposed projects.
The advantages of BPT estimates include:
Simplicity. Users need only know the anticipated or
historical level of emission reductions.
Resource efficiency. Generating benefits estimates
requires only a simple spreadsheet.
Speed. Results can be generated very quickly.
Disadvantages of the BPT estimates include:
Limited ability to account for spatial heterogeneity.
The BPT estimates are best viewed as the average
benefits of emission reductions within a specific
spatial scaleeither nationwide or within one of a
few specific urban areas. In general, the BPT esti-
mates are most appropriate for characterizing the
benefits of broad-scale emission reductions.
Inflexible. Users are unable to modify any of the
assumptions within the BPT metrics, including the
selection of C-R functions, year of population expo-
sure, valuation functions, or air quality modeling.
Based on multiple assumptions. A series of model-
ing assumptions are embedded within the BPT
metrics. Consequently, the greater the divergence
between these embedded assumptions and the
policy context to which the user applies the BPT
metrics, the greater the uncertainty.
A challenge with using BPT measures arises if a clean
air project reduces emissions of multiple pollutants
simultaneously (e.g., SO2 and NO ). In order to reach
a more accurate benefit-per-ton estimate, is important
to apportion the benefits among each of the multiple
types of emission reductions.
Sophisticated Modeling Approaches
Two sophisticated modeling approaches, which vary
in terms of complexity, are used to quantify the hu-
man health impacts of air quality changes: integrated
modeling and linked modeling.
Integrated Modeling
Screening-level integrated models include emissions,
air quality, health effects, and economic valuation
within a single software application that runs quickly
on a desktop computer.
An integrated model typically allows the user to
enter potential emissions from one or more emis-
sion categories, and then apply a series of methods
to estimate air quality changes, population exposure,
avoided health effects, and the economic values of
the quantified benefits. These models are not as rigor-
ous as the linked approach, but can quickly enable a
less experienced analyst to prepare a screening-level
analysis of many different clean energy alternatives.
EPAs COBRA model is an example of an integrated
screening-level model.
Integrated Modeling with COBRA
EPAs Co-Benefits Risk Assessment (COBRA) model is
a computer-based screening model that employs user-
specified emission reduction estimates to estimate air
quality changes and health effects. It is a stand-alone
Windows application that enables users to:
Approximate the impact of emission changes on
ambient air pollution,
Translate these ambient air pollution changes
into related health effect impacts,
Monetize the value of those health effect
impacts, and
Present the results in various maps and tables.
Using COBRA enables policy analysts to quickly and
easily obtain a first-order approximation of the benefits
of different policy scenarios and to compare outcomes
in terms of air quality (i.e., changes in PM concentra-
tions and pollutants associated with the secondary for-
mation of PM, at the county, state, regional, or national
level) or health effects. COBRA is designed to allow
users to quickly and easily analyze the health effects of
changes in emissions of PM.
The COBRA screening tool is based on the following
methodology.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 120
-------
EXAMPLES OF AIR QUALITY HEALTH MODELS
COBRA (a screening-level integrated model)
Suited to less-experienced modelers.
Requires air pollution emissions data, which the model
converts to air quality changes, as an input.
Includes health effects of PM.
Uses EPA-provided default concentration-response (C-R)
functions and economic values.
BenMAP (a linked model)
Suited to experienced modelers, although a new one-step
approach improves accessibility and training is available.
Requires air quality data, which must be estimated
exogenously, as an input.
Includes health effects of PM and ozone.
Uses EPA-provided C-R functions and economic values,
and also allows user-specified functions.
The model contains detailed emission estimates for
the years 2010 and 2015, developed by EPA. Before
running a scenario, users must select one of these
years as the baseline for their scenario.
Users can then create their own scenarios by mak-
ing changes to the emission estimates specified by
the chosen baseline. Changes in PM25, SO2, NOx,
NH3, and VOC emissions can be specified at the
county, state, or national level.
COBRA incorporates user-defined emission
changes into a reduced form air quality model, the
Source Receptor (S-R) Matrix, to estimate the ef-
fects of emission changes on PM concentrations.
COBRA uses concentration-response (C-R) func-
tions to link the estimated changes in PM concen-
trations to a number of health endpoints, including
premature mortality, chronic bronchitis, and asth-
ma. The C-R functions are based on recent epide-
miological studies and are consistent with BenMAP
and recent EPA regulatory impact analyses.
COBRA monetizes the health effects using eco-
nomic value equations based on those approved in
recent EPA rulemakings.
COBRAs use of default C-R function and economic
values for health effects removes the burden of select-
ing these functions and values for users with limited
HEALTH ENDPOINTS INCLUDED IN COBRA
Mortality.
Chronic and acute bronchitis.
Non-fatal heart attacks.
Respiratory or cardiovascular hospital admissions.
Upper and lower respiratory symptom episodes.
Asthma effects, exacerbations, and emergency room visits.
Shortness of breath, wheeze, and cough (in asthmatics).
Minor restricted activity days.
Work loss days
air quality and health modeling experience. The default
values in the model are updated to be consistent with
current EPA benefits methods. However, this strength
in ease of use is also a key limitation because COBRA
cannot incorporate more sophisticated air quality and
health effect modeling techniques, http://epa.gov/state-
localctimate/resources/cobra.html
Linked Modeling
Linked models are rigorous methods that combine
emission estimation, air quality estimates, population
data, baseline health data, and health concentration-
response functions in a geographic-based analysis. This
approach uses a series of separate models in sequence:
a typical sequence of linked models begins with an
electricity generation model, followed by an emissions
model, an air quality model, a health effects model,
and finally an economic valuation model. The results of
each major modeling step is used as an input into the
next, resulting in a rigorous overall analysis relying on
a series of state-of-the-art modeling components.
While such approaches can be data- and resource-
intensive, standard methods and models are available.
Linked health effects modeling translates estimated
changes in air quality into avoidable cases of a wide
range of health effects. EPAs methods and models for
conducting health analysis have been reviewed by EPAs
Science Advisory Board and the National Academy of
Science, and are widely used by EPA, as well as state
and local governments, as a routine part of developing
air quality programs. An example of a linked model for
health effects and valuation is EPAs BenMAP.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 121
-------
HOW ARE STATES USING COBRA?
Connecticut worked with EPA and NESCAUM to quantify the
economic, air quality, and health benefits of policy options
while developing the state's 2005 Climate Change Action Plan.
The COBRA model showed that while "the state's (existing)
energy efficiency program...was known to achieve a $3 to
$1 direct return on investment based on electricity savings...
an additional $4 to $1 payback in terms of reduced health
costs and public health benefits was identified as a result of
reductions in criteria air pollutants."
Source: Connecticut GSC on Climate Change, 2005.
Linked Modeling with BenMAP
EPA's Benefits Mapping and Analysis Program
(BenMAP) is a Windows-based program that enables
users to:
Estimate the health effects for numerous health
endpoints associated with changes in ambient O3
and PM concentrations.
Monetize the value of health effects.
Visually inspect results with maps of air pollution,
population, incidence rates, incidence rate changes,
economic valuations, and other types of data at the
county, state, or national level using geographic
information systems (GIS).
BenMAP systematically analyzes the health and eco-
nomic benefits of air pollution control policy scenarios.
It is designed to provide flexible and timely analysis,
ensure that users can understand the assumptions
underlying the analysis, and adequately characterize
uncertainty and variability. As a first step, BenMAP
estimates impacts to populations from the year f 990
to 2030 according to race, gender, age, and ethnicity.
These data are then used to estimate health impacts
according to sub-population.
The BenMAP modeling approach is illustrated in Fig-
ure 4.2.4 and described below.
BenMAP applies the damage function approach,
a technique used to estimate the health impacts
resulting from changes in air pollution. The damage
function incorporates air pollution monitoring data,
air quality modeling data, Census data, population
projections, and baseline health information to re-
late a change in ambient concentration of a pollutant
FIGURE 4.2.4 BENMAP HEALTH
IMPACTS
U.S. Census
Data
Air Quality
Modeling
Health
Functions
A BenMAP
MODELING PROCEDURE
Population
Estimates
1
+ 1^ Population L
P Exposure
1
Adverse
Health Effects
^
Economic
Costs
Input
Population ^
Projections
Air Quality + 1
Monitoring
Baseline
Incidence Rates
Validation ^ I
Functions
+ User Input Choice
Result from Input
to population exposure, and quantifies the incidence
of new or avoided adverse health endpoints.
Users typically run BenMAP to estimate the health
impacts of a policy scenario, specifying both base-
line and post-policy air quality levels. BenMAP
then estimates the changes in population exposure.
Air quality information for the baseline and
scenario runs need to be generated exogenously,
either from monitor-based air quality data, model-
based air quality data, or both.19 BenMAP includes
monitoring data for O3, PM, NO2, and SO2 for a
number of years.
BenMAP then calculates the changes in health effect
incidence associated with the change in population
exposure by using concentration-response functions
(C-R) derived from the epidemiological literature
and pooling methods specified by the user.20 Ben-
MAP uses the estimate of statistical error associated
with each C-R function to generate distributions of
19 BenMAP accepts air quality output from a variety of models, including
Regulatory Model System for Aerosols and Deposition (REMSAD), the Com-
prehensive Air Quality Model with Extensions (CAMx), the Urban Airshed
Monitoring-Variable grid model (UAM-V), the Community Multi-Scale Air
Quality Model (CMAQ) and EPA's Response Surface Model (RSM). BenMAP
can also accept other model results by changing the default input structure.
20 Pooling is a method of combining multiple health effects estimates to gener-
ate a more robust single estimate of health impacts.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 122
-------
incidence estimates, as well as a central point esti-
mate. These distributions are helpful for characteriz-
ing the uncertainty associated with this component
of the health impact assessment.
BenMAP also calculates the economic value of the
avoided or incurred health effects based on valua-
tion approaches from the published economics lit-
erature. The estimated economic value of an avoid-
ed health outcome is multiplied by total change
in events to determine the health benefits of air
quality improvements. As with the C-R functions
described above, the valuation functions include
estimates of statistical error that BenMAP uses to
generate distributions of results (EPA, 2003).
One of BenMAP s strengths is that it includes large da-
tabases of C-R functions and economic valuations from
which the user can select when performing an analysis.
Users can also add new functions. In addition, by using
air quality modeling data or actual monitoring data, it
provides robust estimates of health impacts with a high
degree of spatial resolution (Davidson et al., 2003).
http://www.epa.gov/air/benmap/
Key Considerations When Selecting Methods
to Estimate Health Effects and Associated
Economic Impacts of Clean Energy
The following issues can be considered when selecting
a basic or sophisticated modeling approach:
Pollutants to be analyzed. While health modeling
for O3 and PM is the most common approach,
analyses are also conducted for SO2 emissions, CO,
Hg, and other air toxics emitted by conventional
electricity generation.
Selection of health effects. Even though a long list of
health effects analysis is possible, in some circum-
stances a significantly smaller set may be sufficient.
EPA has quantified PM-related health effects in-
cluding premature mortality in adults and infants,
chronic bronchitis, non-fatal heart attacks, hospital
admissions for respiratory and cardiovascular
diseases, emergency room treatment for asthma,
asthma attacks, and various "symptom-days" (in-
cluding work loss days). Quantified ozone-related
health effects include respiratory hospital admis-
sions and emergency room visits, and "symptom-
days" (including school absences). Recent health
research indicates that O3 is also associated with
premature mortality, which has been included as a
new health effect in recent EPA analyses.
HOW BENMAP HAS BEEN USED IN CLEAN ENERGY ANALYSIS
For testimony to the Minnesota Public Utilities Commission
about building a new clean energy electricity generating facility.
Excelsior Energy compared the air quality and health effects of
two proposed 600 MW integrated gasification and combined
cycle (IGCC) units with two comparable supercritical pulverized
coal (SCPC) units. The analysis used REMSAD to model Hg and
PM air quality changes, and BenMAP to estimate and value the
PM-related health effects. For the IGCC option, for example,
the study found that installing IGCC technology would reduce
annual emissions by 2,600 tons of SO2, 600 tons of NOx, and
12 pounds of Hg. The largest impacts on PM2.5 concentrations
occurred within 80 km of the proposed facility, although
small PM impacts also occurred hundreds of miles downwind,
affecting millions of additional people. The analysis also found
that in 2012, the IGCC units would avoid 12 premature deaths
nationally, 20 heart attacks (infarctions), eight new cases of
chronic bronchitis, and 200,000 work loss days, and quantified
estimates of other health effects ranging from hospital
admissions to asthma attacks. The annual value of the one
year of reduced health effects was estimated to be $99 million
nationally, with $24 million occurring within Minnesota.
Sources: Excelsior Energy, 2005.
Selection of C-R functions for health analysis. The
specific mathematical functions that estimate the
changes in health effects from changes in ambient
air quality are typically derived from epidemiologi-
cal research. For most of the health effects selected
for an analysis, a variety of alternative C-R functions
are available from different sources. It is important
to carefully select functions that appropriately
reflect the central estimates and the range of diverse
results from different published health studies,
while striving to avoid double counting and mini-
mizing the omission of important health effects.
Time span. Estimating the health effects for differ-
ent pollutants requires different time spans. Ozone
health effects typically require hourly air quality
estimates, but analysis is sometimes limited to the
ozone season, or even modeling a one or two week
episode during the peak ozone period. Estimating
the health effects of PM, on the other hand, typi-
cally requires daily air quality estimates throughout
the entire year, or estimates of the impact on the
annual mean PM level.
Geographic scope. Every health effects estimation
procedure operates at some level of geographic res-
olution. Some health effects models use the county
level for the analysis, while others match the level
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 123
-------
of the air quality model and use a rectangular grid
system. (Hubbell, 2008)
Selection of methods for estimating the economic
value of avoided health effects. Estimating the
economic value of the avoided cases of each health
effect allows stakeholders to more directly compare
the economic benefits of a clean energy project
with the project's costs. Economic values for each
health effect are derived from economic literature,
and must be carefully matched to the types of
avoided health effects estimated in an analysis.
4.3 CASE STUDIES
4.3.1 TEXAS EMISSIONS REDUCTION
PLAN (TERP)
Benefits Assessed in Analysis
NO reductions
x
Clean Energy Program Description
In 2001, the 77th Texas Legislature established the Texas
Emissions Reduction Plan (TERP) with the enactment
of Senate Bill 5, which required the Texas Commission
on Environmental Quality (TCEQ) to promote EE/RE
to meet ambient air quality standards and to develop
a methodology for computing emission reductions for
State Implementation Plans (Haberl et al., 2004). To
improve Texas air quality, TERP adopted the goal of
implementing cost-effective EE/RE measures to reduce
electric consumption by 5 percent per year for five
years, beginning in 2002, using a variety of mandatory
programs and voluntary financial incentive programs
in non-attainment and affected counties.
These programs included:
Texas Building Energy Performance Standards for
residential and commercial building construction.
An emissions reduction incentive grants program,
which provides grants to offset costs associated
with reducing NOx emissions.
A new technology research and development pro-
gram, which provides incentives to support R&D
that will reduce pollution in Texas.
A small business program, which helps small busi-
nesses and others participate in the TCEQs incen-
tive program.
Methods Used
To meet annual reporting requirements, the TCEQ
worked with the State Energy Conservation Office
(SECO), the Public Utility Commission (PUC), the
Energy Systems Laboratory (ESL) and the Electric Reli-
ability Council of Texas (ERGOT) to develop method-
ologies for quantifying the NOx emission reductions
associated with energy savings from TERP clean en-
ergy projects. A key step in that process was to develop
uniform accounting procedures to be applied to the
energy savings across the different programs. For ex-
ample, during 2001 and 2002, NO emission reduction
values could not be integrated across programs because
they were reported to the TCEQ by several agencies in
disparate units (i.e., lbs-NOx/year vs. tons-NOx/OSD),
time frames (i.e., annual, average daily), and variations
in conversion factors (i.e., lbs-NOx/MMBtu, g-NOx/
kilojoule, tons-NOx/MWh).
Each reporting agency used a unique methodology
to estimate energy savings from its programs, all of
which were subsequently converted to NO emission
reductions using eGRID average emissions factors as
described below.
For SECO, Energy Service Companies (ESCOs)
reported stipulated energy savings for about 100
projects to SECO. These annual estimates of energy
savings were then converted into average daily sav-
ings for use in the NO emissions calculations for
t> X
the Ozone-Season-Day (OSD) using eGRID.
For the PUC's utility-based programs, calculated
annual savings for more than 100,000 projects are
reported to the PUC using a standard template.
These savings are then converted to average daily
OSD savings for use in the NOx emissions calcula-
tions for the OSD using eGRID*
For code-compliant construction programs, the
ESL developed simulation models for residential
buildings using the DOE-2.1e simulation program.
ESLs models were then linked to eGRID to auto-
matically convert energy savings into NOx emis-
sion reductions.
For green power programs, 15-minute metered
data, obtained from ERGOT, and average daily
values for the Ozone Season Period were used to
represent the OSD electricity and NOx reductions
using eGRID.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 124
-------
Results
The 2007 annual report on energy savings and
emission reductions for energy-code- compliant
new residential single, multi-family, and commer-
cial construction reported the following findings
(Haberl et al, 2007):
> The annual energy savings in 2006 amounted to
498,582 megawatt hours (MWh) of electricity
and 576,680 million BTUs of natural gas, which
led to 361 tons of NOx reductions in 2006.
On a peak summer daywhen ozone forma-
tion is at its worstthe NOx reductions in
2006 were calculated to be 2.23 tons per day.
Cumulative NOx reductions, projected to 2013,
from energy efficiency savings from code-
compliant new residential and commercial
construction were determined to be 2,121
tons/year and 10.75 tons/peak-day.
For More Information
Energy Efficiency/Renewable Energy Impact in the
Texas Emissions Reduction Plan (TERP): Volume 1 -
Summary Report. Prepared for the Texas Commis-
sion on Environmental Quality (TCEQ). August
2007, revised December 2007. http://esl.eslwin.
tamu.edu/docs/documents/ESL-TR-07-12-01.pdf
Texas Commission on Environmental Quality
(TCEQ). http://www.tceq.state.tx.us/
* Texas A&M University, Energy Systems Labora-
tory, Senate Bill 5. http://esl.eslwin.tamu.edu/
senate-bill-5.html
4.3.2 WISCONSIN - FOCUS ON ENERGY
PROGRAM
Benefits Assessed
Energy savings
Renewable energy generation
Reductions of NOX
Reductions of CO2
Reductions of SOX
Reductions of mercury
Energy bill savings
Clean Energy Program Description
Funded by the Utility Public Benefits fund created by
the Wisconsin State Legislature in 1999, the Wisconsin
Focus on Energy Program aims to reduce energy use
and advance clean energy supplies throughout Wiscon-
sin by:
Promoting energy efficient practices and equipment
in new and existing buildings across the residential,
industrial, commercial, agricultural, and govern-
ment sectors;
Promoting the installation of renewable energy;
Educating the public about renewable energy; and
Providing grants for research on the environmental
impacts of electric generation.
Focus on Energy programs include the Wisconsin
ENERGY STAR Products (ESP) program, Wisconsin
ENERGY STAR Homes (WESH), Home Performance
with ENERGY STAR (HPWES), as well as other sector-
and renewable-energy-focused programs (DOA, 2005).
Methods Used
To analyze how efficiency programs affect air emis-
sions, the Wisconsin DOA enlisted an independent
program evaluation contractor to comprehensively
analyze the emission impacts of the states efficiency
programs by quantifying emission reductions for dif-
ferent seasons and hours of the day.
The general approach DOE used to estimate emissions
from clean energy programs was to:
Develop seasonal and off-peak emissions factors
expressed in pounds of pollutant per MWh or
GWh for nitrogen oxides (NOx), sulfur dioxides
(SOx), carbon dioxide (CO2), and mercury (Hg) for
the regional electricity supply system serving Wis-
consin. The DOA used EPA continuous emission
monitoring data on historical plant operations and
emissions to estimate which generating plants were
"on the margin" during different time periods.21
Multiply the emissions factors by the energy savings
from Focus on Energy programs efforts to produce
an estimate of the total avoided emissions.
21 EPA Office of Air and Radiation. "Acid Rain/OTC Program Hourly Emis-
sions Data." http://www.epa.gov/airmarkets/emissions/raw/index.html
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 125
-------
To determine when the energy savings occurred so that
it could apply the corresponding emissions factor (e.g.,
seasonal, hourly), DOA divided the annual energy
savings for each measure into four bins: winter peak,
winter off-peak, summer peak, and summer off-peak.
DOA made these determinations based on internal
evaluations of the operating characteristics of its pro-
grams, along with work done by the New Jersey Clean
Energy Collaborative and reported in Protocols to Mea-
sure Resource Savings, http://www.njcleanenergy.com/
filesffilefProtocols_REVISED_VERSION_l.pdf
These calculations assume that the energy savings
result in reduced generation at the power plants
that are operating on the margin during a par-
ticular time of day or season. As described earlier
in this chapter, the marginal generator is the last
generator called upon to meet current demand
for electricity, and it can vary over time (within a
day and across seasons) as demand changes. Using
emissions factors to estimate avoided emissions
also assumes that reduced demand is perfectly cor-
related with reduced emissions.22
Results
The emission benefits for Focus on Energy's business
and residential programs by peak/season and program
22 This may not always be true. For example, even if demand is reduced in
Wisconsin, Wisconsin generators may continue operating as they did before
and sell more power out of state.
from July 1, 2001 through September 30, 2003 are sum-
marized in Table 4.3.1.
Based on a more recent study update published in 2006,
DOA estimates that from July 1, 2001 through June
30, 2006, its programs saved nearly 1 billion kWhs and
nearly 50 million therms in annual energy consump-
tion. This is equivalent to annual energy savings of
almost $80 million for electricity (kWh) and nearly $50
million in gas savings (therms), and a lifetime dollar
value of energy costs saved totaling more than $660
million for electricity saved and more than $430 million
for gas saved. These programs have displaced annual
emissions from power plants and utility customers by:
5.8 million pounds of NOX,
2.6 billion pounds of CO2,
11.4 million pounds of SOX, and
46 pounds of mercury.
With stable funding over the next ten years, the state
projects that the Focus on Energy program will add
nearly $1 billion in value to Wisconsin's gross state
product (DOA, 2006).
Performing this comprehensive emissions factor deriva-
tion improved the accuracy of avoided emission esti-
mates from Focus on Energy efficiency programs and
allowed the program to take into account differences
TABLE 4.3.1 EMISSION REDUCTIONS FROM FOCUS ON ENERGY BUSINESS AND RESIDENTIAL
PROGRAMS BY PEAK AND SEASON PERIODS (JULY 1, 2001 - SEPTEMBER 30, 2003)
Summer Off-peak
Business Programs
Residential Programs
Pounds
444,544
216,265
89,429,423
2.1
300,946
146,406
60,541,736
1.4
Summer Peak
473,349
222,184
86,362,026
1.7
311,951
146,426
56,915,134
1.1
Winter Off-peak
715,544
286,218
112,858,634
2.6
597,750
239,100
94,279,589
2.2
Winter Peak
863,768
366,635
125,961,032
2.7
681,608
289,316
99,397,104
2.1
On-site Natural Gas
757
126,146
151,313,733
Total
2,497,206
1,091,302
414,611,115
9.1
1,892,255
821,248
311,133,562
6.8
Source:Erickson ef a/., 2004.
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 126
-------
across energy efficiency measures in terms of the
distribution of energy savings over sectors and periods
of time, and develop an optimal portfolio of energy ef-
ficiency programs with respect to emission reductions.
Using this type of approach, program designers can use
the seasonal and peak emissions factors combined with
information on load patterns for various types of equip-
ment and businesses to target program efforts towards
those areas that would produce the most emissions re-
ductions for a given level of effort (Erickson et al., 2004).
For More Information
Estimating Seasonal and Peak Environmental
Emission Factors - Final Report. Prepared by PA
Government Services for the Wisconsin DOA.
May 2004. http://www.doa.state.wi.us/docs_view2.
asp?docid=2404
Focus on Energy Public Benefits Evaluation - Semi-
annual Summary Report. Prepared by PA Govern-
ment Services for the Wisconsin DOA. September
14,2005. http://www.doa.state.wi.us/docs_view2.
asp?docid=5237
Focus on Energy Public Benefits Evaluation - Semi-
annual Summary Report. Prepared by PA Govern-
ment Services for the Wisconsin DOA. September
27,2006. http://www.focusonenergy.com/files/
Document_Management_System/Evaluation/
semiannualyearendfy06_evaluationreport.pdf
Focus on Energy Program http://www.focusonen-
ergy.com/
Developing a Baseline Emissions Profile
URL Address
DOE's State Energy Consumption, Price, and Expenditure Estimates (SEDS) database.
http://www.eia.doe.gov/emeu/states/_seds.html
The ICLEI Cities for Climate Protection program Web site has greenhouse gas
emissions inventories and plans developed by many major cities in the United States.
h ttp://www. icleiusa. org/action-cen ter/learn -
from-others/action-plans-inventories
State Energy Offices often have energy use data and projections. For example, the
New York State Energy Research and Development Authority (NYSERDA) published
such data in "Patterns and Trends: New York State Energy Profiles (1993-2007)" (2009).
http://www.nyserda.org/energy_information/
patterns%20&%20trends%201993-2007.pdf
Basic Modeling Methods
Defining operating characteristics/data on load profiles
The California Database for Energy Efficient Resources (DEER), sponsored by the
California Energy Commission and California Public Utilities Commission (CPUC),
provides estimates of energy and peak demand savings values, costs, and effective
useful life of efficiency measures.
h ttp://www. energy, ca.gov/deer/
NREL's HOMER simplifies the task of evaluating the economic and technical feasibility
of design options for remote, stand-alone, and distributed generation applications
(both off-grid and on-grid).
h ttp://www.nrel.gov/homer/
National Assessment of Emissions Reduction of Photovoltaic (PV) Power Systems
(Analysis Group for Regional Electricity Alternatives, Laboratory for Energy and the
Environment, and the Massachusetts Institute of Technology, 2004).
http://www.masstech.org/IS/public_policy/dg/
resources/2004_PV-Avoided-Emissons_Main-
Rept_MlT-Conners-et-al-l.pdf
Some states or regions have technology production profiles in efficiency and
renewable energy potential studies, e.g.. Energy Efficiency and Renewable Energy
Resource Development Potential in New York State: Volume Four contains energy
production by costing period for some renewable resources (New York State Energy
Research and Development Authority, 2003).
http://www.nyserda.org/Energy_lnformation/
energy_state_plan.asp
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 127
-------
The appendices to the Connecticut Energy Conservation Management Board's
Maximum Achievable Potential Study (2004) have detailed data about efficiency
measures, including kW in summer and winter (available upon request).
URL Address
http://www.ctsavesenergy.org/ecmfa/
documents. php?section=30
NREL's PV Watts calculates location-specific monthly energy production (kWh) from
photovoltaic systems.
http://www.nre/.go\//rredc/p\/watts/
Data on emissions rates and capacity factors
EPA's eGRID database provides information on emissions by individual power plants,
generating companies, states, and regions of the power grid.
http://www.epa.gov/egrid
The NEPOOL Marginal Emission Rate Analysis report provides marginal emission rates
during four time periods (ozone/one-ozone and peak/off-peak) for NOx, SOx, CO2 for
the NEPOOL region.
http://www./so-ne.com/genrt/on_resrcs/
reports/emission/in dex. h tm I
The Emission Reduction Workbook (OTC Workbook) (Keith, G., D. White, and B.
Biewald, 2002) was developed for the Ozone Transport Commission in 2002.
http://www.synapse-energy.com/Down/oads/
SynapseReport.2002-12.OTC.OTC-Emission-
Reduction-Workbook-v-2.1.02-34-Workbook.xls
EPA's Acid Rain data (recently moved to the Clean Air Markets website) provides
hourly data on SO2, NOx, and CO2 emissions for Acid Rain and NOx SIP Call/OTC units
since 1997 (since 1995 for coal-fired units).
http://camddataandmaps.epa.gov/gdm/index.
cfm ?fuseaction=prepackaged.select
Electric Energy Efficiency and Renewable Energy in New England: An Assessmen t of
Existing Policies and Prospects for the Future (the Regulatory Assistance Project and
Synapse Energy Economics, 2005) describes an analysis that used the OTC workbook
to estimate emissions reductions from efficiency and renewables in New England.
http://www.synapse-energy.com/Down/oads/
SynapseReport.2005-05.RAP-EPA.Efficiency-
and-Renewable-Energy-in-New-England.04-23.
pdf
Emerging Tools for Assessing Air Pollutant Emission Reductions from Energy
Efficiency and Clean Energy: Phase II Final Report. Global Environment & Technology
Foundation, January 31 2005.
h ttp://www.4c/eana/r.org/
EmissionsModelingPhasellFinal.pdf
Model Energy Efficiency Program Impact Evaluation Guide provides guidance on
model approaches for calculating energy, demand, and emissions savings resulting
from energy efficiency programs. The Guide is provided to assist in the implementation
of the National Action Plan for Energy Efficiency's five key policy recommendations
and its Vision of achieving all cost-effective energy efficiency by 2025.
http://www.epa.gov/cleanrgy/documents/
evaluation_guide.pdf
Using Electric System Operating Margins and Build Margins: Quantification of Carbon
Emission Reductions Attributable to Grid Connected COM Projects (Biewald, B. 2005),
prepared for the United Nations Framework Convention on Climate Change (UNFCC),
analyzed the impact of reductions in electricity demand and renewable generation on
http://www.synapse-energy.com/Down/oads/
SynapseReport.2005-09.UNFCCC.Using-
Electric-System-Operating-Margins-and-Build-
Margins-.05-031.pdf
CO2 emissions.
Methods for Estimating Emissions Avoided by Renewable Energy and Energy
Efficiency (Keith, G. and B. Biewald, 2005), prepared for U.S. Environmental Protection
Agency, evaluates several methods of estimating displaced emissions without using a
dispatch model.
http://www.synapse-energy.com/Down/oads/
SynapseReport.2005-07.PQA-EPA.Displaced-
Emissions-Renewables-and-Efficiency-
EPA.04-55.pdf
Modeling Demand Response and Air Emissions in New England (Keith, G., B. Biewald,
D. White, and M. Drunsic, 2003), prepared for the U.S. Environmental Protection
Agency, presents an analysis of the impact of reductions in electricity demand and
renewable generation on air emissions.
http://www.synapse-energy.com/Down/oads/
SynapseReport.2003-09.US-EPA.NE-DR-and-
AE-Modeling.03-01.pdf
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 128
-------
URL Address
Sophisticated Modeling Methods
Electric Dispatch Models
Electric dispatch models that can be used to assess displaced emissions include:
GE-MAPS (Multi-Area Production Simulation)
Market Analytics (PROSYM)
PROMOD IV
GE-MAPS
http://www.gepower.com/prod_serv/products/
utility_software/en/ge_maps/index.htm
Market Analytics
h ftp://www.ven tyx.com/analytics/market-
analytics.asp
PROMOD IV
h ttp://www.ven tyx.com/analytics/promod.asp
Capacity Expansion Models
Energy Portfolio Management: Tools and Practices for State Public Utility
Commissions (Steinhurst, W., D. White, A. Roschelle, A. Napoleon, R. Hornby, and B.
Biewald, 2006) describes a sample of capacity expansion models.
http://www.synapse-energy.com/Downloads/
SynapseReport.2006-07.NARUC.Portfolio-
Management-Tools-and-Practices-for-
Regula tors. 05- 042.pdf
The Hudson River Foundation financed the Clean Electricity Strategy for the Hudson
River Valley (Synapse Energy Economics and Pace Law School Energy Project, 2003).
This report explores the air-emissions reductions that would likely result from the
implementation of a proposed clean energy plan, consisting of new energy efficiency
programs, renewable generation, combined heat and power, and retrofit projects.
http://www.synapse-energy.com/Downloads/
SynapseReport.2003-10.Pace.Hudson-River-
Clean -Energy-Stra tegy. 02- 23.pdf
Capacity expansion models that can be used to assess displaced emissions include:
Integrated Planning Model (IPM) (ICF International)
National Energy Modeling System (NEMS) (U.S. DOE)
ENERGY 2020
Integrated Planning Model (IPM)
http://www.icfi.com/Markets/Energy/energy-
modeling.asp#2
NEMS
http://www.eia.doe.gov/oiaf/aeo/overview/
index.html
ENERGY 2020
h ttp://www. energy2020. com/
Quantifying Air Quality and/or Health Impacts
SCRAM
http://www.epa.gov/ttn/scram/
REMSAD
h ftp://remsad.sain tl. com
CAMx
h ttp://www. camx. com
UAM-V
http://uamv.saintl.com
CMAQ
http://www.epa.gov/AMD/CMAQ/
CM A Qscien ceDoc.html
CALPUFFandAERMOD
http://www.epa.gov/scram001/dispersion_
prefrec.htm
COBRA
http://epa.gov/statelocalclimate/resources/
cobra.html
BenMAP
http://www.epa.gov/air/benmap/
ASAP
http://www.epa.gov/ttn/ecas/asap.html
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 129
-------
URL Address
Texas Case Study
Energy Efficiency/Renewable Energy Impact in the Texas Emissions Reduction
Plan (TERP): Volume 1 - Summary Report. Prepared for the Texas Commission on
Environmental Quality (TCEQ). August 2007, revised December 2007.
http://esl.eslwin.tamu.edu/docs/documents/
ESL-TR-07-12-01.pdf
Texas Commission on Environmental Quality (TCEQ).
http://www.tceq.state.tx.us/
Texas A&M University, Energy Systems Laboratory, Senate Bill 5.
http://esl.eslwin.tamu.edu/senate-bill-5.html
Wisconsin Case Study
Estimating Seasonal and Peak Environmental Emission Factors - Final Report.
Prepared by PA Government Services for the Wisconsin DOA. May 2004.
http://www.doa.state.w/.us/docs_w'ew2.
asp?docid=2404
Department of Administration, State of Wisconsin. 2005. Focus on Energy Public
Benefits Evaluation - Semiannual Summary Report. Prepared by PA Government
Services for the Wisconsin DOA. September.
http://www.doa.state.w/.us/docs_w'ew2.
asp?doc/d=5237
Department of Administration, State of Wisconsin. 2006. Focus on Energy Public
Benefits Evaluation - Semiannual Summary Report. Prepared by PA Government
Services for the Wisconsin DOA. September.
http://www.focusonenergy.com/ftVes/
Document_Management_System/Evaluation/
semiannualyearendfy06_evaluationreport.pdf
Focus on Energy Program
h ftp ://www. focusonenergy. com/
New Jersey Clean Energy Program. 2007. Protocols to Measure Resource Savings,
December.
http://www.njcleanenergy.com/files/file/
Protocols_Final_12-20-07_%5Bl%5D.pdf
EPA Office of Air and Radiation. "Acid Rain/OTC Program Hourly Emissions Data."
http://camddataandmaps.epa.gov/gdm/index.
cfm ?fuseaction=prepackaged.select
References
URL Address
Connecticut GSC on Climate Change. 2005. CCCAP. GSC on Climate Change.
Connecticut Climate Change Web site. State Action Plan.
Davidson, K., A. Hallberg, D. McCubbin, and B. Hubbell. 2003. Analysis of PM2.5 Using
the Environmental Benefits Mapping and Analysis Program (BenMAP). Presented at
the 2nd AirNet Annual Conference/NERAM International Colloquium. November.
Department of Administration, State of Wisconsin. 2005. Focus on Energy Public
Benefits Evaluation - Semiannual Summary Report. Prepared by PA Government
Services for the Wisconsin DOA. September.
Department of Administration, State of Wisconsin. 2006. Focus on Energy
Public Benefits Evaluation - Semiannual Summary Report FY06. Prepared by PA
Government Services for the Wisconsin DOA. September.
Erickson et al. 2004. Erickson, J., C. Best, D. Sumi, B. Ward, B. Zent, and K. Hausker.
Estimating Seasonal and Peak Environmental Emission Factors - Final Report.
Prepared by PA Government Services for the Wisconsin DOA. May.
Excelsior Energy. 2005. Air Quality and Health Benefits Modeling: Relative Benefits
Derived from Operation of the MEP-I/II IGCC Power Station. December.
Fann, N, Fulcher, C, Hubbell, B. The Influence of Location, Source, and Emission Type in
Estimates of the Human Health Benefits of Reducing a Ton of Air Pollution. Submitted
for publication October 2008, Journal of Air Quality, Atmosphere and Health.
h ttp://www. ctclima techange. com/
StateActionPlan.html
h ttp://www. informaworld. com/smpp/con ten t~
db=all?content=10.1080/15287390600884982
(fee for full text)
http://www.doa.state.w/.us/docs_w'ew2.
asp?doc/d=52J7
http://www.focusonenergy.com/ftVes/
Document_Management_System/ฃVa(uat/on/
semiannualyearendfy06_evaluationreport.pdf
http://www.doa.state.wi. us/docs_view2.
asp?docid=2404
http://mncoalgasplant.com/15%20Exhibit%20D.
pdf
http://www.springerlink.com/content/
bl 6jx57531 877j31/fulltext.pdf
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 130
-------
References
Haberl, J., C. Gulp, B. Yazdani, D. Gilman, T. Fitzpatrick, S. Muns, M. Verdict, M.
Ahmed, Z. Liu, J. Baltazar-Cervantes, C. Montgomery, K. McKelvey, J. Mukhopadhyay,
and L. Degelman. 2007. Energy Efficiency/Renewable Energy Impact in the Texas
Emissions Reduction Plan (TERP): Volume 1 - Summary Report. Prepared for the
Texas Commission on Environmental Quality (TCEQ). August, revised in December.
Hubbell, Bryan, Fann, N, Levy, J, The Methodological Considerations in Developing
Local-scale Health Impact Assessments: Balancing National, Regional, and Local
Data. Submitted for publication October 2008, Journal of Air Quality, Atmosphere
and Health.
Imhoff, B. 2006. Assessment of Environmental Benefits (AEB) Modeling System.
Presented at the EPA Air Innovations Conference, Denver. September 6.
IPCC. 2008. Intergovernmental Panel on Climate Change Web Site: Methodology
Reports.
Keith, G. and B. Biewald. 2005. Methods for Estimating Emissions Avoided by
Renewable Energy and Energy Efficiency. July 8.
Metropolitan Washington Council of Governments. 2007. State Implementation Plan:
Plan to Improve Air Quality in the Washington, DC-MD-VA Region, May 23. Table 609,
p.6-62
National Action Plan for Energy Efficiency. 2007. Model Energy Efficiency Program
Impact Evaluation Guide. Prepared by Steven R. Schiller, Schiller Consulting, Inc.
November.
NERC. 2009. North American Electric Reliability Corporation Web Site: Key Players -
Regions.
New Mexico Climate Change Advisory Group. 2006. Climate Change Action Plan.
Final Report. December.
NYSERDA. 2009. New York State Energy Research and Development Authority,
Patterns and Trends: New York State Energy Profiles (1993-2007). December.
Synapse. 2003. Synapse Energy Economics and Pace Law School Energy Project.
Texas Commission on Environmental Quality. 2008. Sources of Air Pollution.
U.S. Census Bureau. 2008. Population Estimates.
U.S. Department of Commerce. 2008. Bureau of Economic Analysis. Regional
Economic Accounts.
U.S. DOE. 2008a. Energy Information Administration. Consumption, Price, and
Expenditure Estimates. State Energy Data System (SEDS). November.
U.S. DOE. 2008b. Energy Information Administration. Environment.
U.S. EPA. 2003. BenMAP User's Manual. Office of Air Quality Planning and Standards.
URL Address
h ttp://esi eslwin.tam u. edu/docs/documen ts/
ESL-TR-07-12-01.pdf
http://www.springerlink.com/content/
bl 6jx57531 877j31/fuUtext.pdf
h ttp://www. cleanairinfo. com/
airinnovations/2006/presentations/Wednesday/
Concurrent%201A/AEB%2015%20Air%20
lnnovations%202006%20Jesse%20Onealpdf
h ttp://www. /pec. ch/publica tions_and_ da ta/
publications_and_data_reports.htm#4
http://www.synapse-energy.com/Downloads/
SynapseReport.2005-07.PQA-EPA.Displaced-
Emissions-Renewables-and-Efficiency-
EPA.04-55.pdf
h ttp://www. m wcog. org/uploads/p ub-
documents/9FhcXg20070525084306.pdf
http://www.epa.gov/cleanrgy/documents/
evaluation_guide.pdf
http://www.nerc.com/page.php?cid=l\9\119
h ttp://www.nmclima techange. us
http://www.nyserda.org/energy_information/
patterns%206f%20trends%201993-2007.pdf
http://www.synapse-energy.com/cgi-bin/
synapsePublications.pl?filter_type=Client&filter_
option=Pace+Law+School+Energy+Project&ad
vanced=false
h ttp://www. tceq.s ta te. tx. us/implemen ta tion/air/
areasource/Sources_of_Air_Pollution.html#top
http://www.census.gov/popest/estimates.php
http://www.bea.gov/regional/index.htm
http://www.eia.doe.gov/emeu/states/_seds.html
h ttp://www. eia. doe.gov/environmen thtml
http://www.epa.gov/air/benmap/
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 131
-------
References
U.S. EPA. 2004. Guidance On State Implementation Plan (SIP) Credits For Emission
Reductions From Electric-Sector Energy Efficiency And Renewable Energy Measures.
August.
U.S. EPA, 2006. Clean Energy-Environment Guide to Action: Policies, Best Practices,
and Action Steps for States. April.
U.S. EPA. 2006a. Technical Support Document for the Proposed PM NAAQS Rule:
Response Surface Modeling. Office of Air Quality Planning and Standards, Research
Triangle Park, NC. February.
U.S. EPA. 2006b. User's Manual for the Co-Benefits Risk Assessment (COBRA) Model.
Developed by Abt Associates for the Climate Protection Partnerships Division, Clean
Energy-Environment State Partnership Program. Washington, DC.
U.S. EPA, 2007. Technology Transfer Network Clearinghouse for Inventories &
Emissions Factors Web Site: Emission Inventory Improvement Program.
U.S. EPA. 2008. Air Pollution Emissions Overview.
U.S. EPA. 2008a. Climate Change - Science Web Site.
U.S. EPA. 2008b. Inventory Of U.S. Greenhouse Gas Emissions And Sinks: 1990-2006.
April. USEPA#430-R-08-005
U.S. EPA. 2008c. Technology Transfer Network
Support Center for Regulatory Atmospheric Modeling: Air Quality Models
U.S. EPA. 2008d. What Are the Six Common Air Pollutants?
U.S. EPA. 2009. Report on the Environment - Outdoor Air.
URL Address
http://www.epa.gov/ttn/oarpg/tl/memoranda/
ereseerem_gd.pdf
http://www.epa.gov/statelocalclimate/
resources/action-guide.html
http://www.epa.gov/scram001/reports/
pmnaaqs_ tsd_rsm_all_021606.pdf
http://www.epa.gov/statelocalclimate/
resources/cobra. html
http://www.epa.gov/ttn/chief/eiip/techreport
http://www.epa.gov/air/oaqps/emissns.html
h ttp://www. epa.gov/clima techange/science/
index.html
h ttp://www. epa.gov/clima techange/emissions/
downloads/08_CR.pdf
http://www.epa.gov/scram001/aqmindex.htm
http://www.epa.gov/air/urbanair/6poll.html
http://cfpub.epa.gov/eroe/index.
cfm?fuseaction=list.listBySubTopic&lv=list.
UstByChapter&ch=466-s=341
CHAPTER 4 | Assessing the Multiple Benefits of Clean Energy 132
-------
CHAPTER FIVE
Assessing the Economic Benefits
of Clean Energy Initiatives
Given the strong link between economic performance
and energy use, it is important for states to account for
the macroeconomic effects of potential clean energy
policies and programs during the process of selecting
and designing these policies. Many studies have shown
that when a state makes cost-effective investments
in energy efficiency and renewable energy, the state's
entire economy will benefit. For example, Wisconsin's
Focus on Energy Program was created to manage ris-
ing energy costs, promote in-state economic develop-
ment, protect the environment, and control the state's
growing demand for electricity. An analysis conducted
by the Wisconsin Department of Administration an-
ticipates that it will meet these objectives while creating
more than 60,000 job years, generating more than
eight billion dollars in sales for Wisconsin businesses,
increasing value added or gross state product by more
than five billion dollars, and increasing disposable
income for residents by more than four billion dollars
between 2002 and 2026 (Wisconsin Department of
Administration, 2007; see text box States Quantify-
ing the Economic Benefits of Clean Energy Policies).
These results demonstrate that positive results from
clean energy investments have spread to the broader
community.
States can estimate the potential economic benefits of
clean energy policies and programs they are consider-
ing by projecting potential changes in the flow of
goods, services, and income within a regional, state,
or local economy. These changes can result in benefits
to key macroeconomic indicators, including employ-
ment, gross state product, economic output, economic
growth, and personal income/earnings. By assessing the
benefits of clean energy on these indicators, states can:
Demonstrate how clean energy can help achieve
economic development goals;
U
o
Q
0 CHAPTER ONE
Introduction
<> CHAPTER TWO
Potential Energy Impacts of Clean Energy
CHAPTER THREE
Electric System Benefits of Clean Energy
CHAPTER FOUR
Air Quality Benefits of Clean Energy
CHAPTER FIVE
Economic Benefits of Clean Energy
APPENDIX A
Catalogue of Clean Energy Case Studies
6 APPENDIX B
Tools and Models Referenced in Each Chapter
CHAPTER FIVE CONTENTS
5.1 How Clean Energy Initiatives Create Macroeconomic
Benefits 134
5.1.1 What are the Direct Effects of Demand-Side
Initiatives? 135
5.1.2 What are the Direct Effects of Supply-Side
Initiatives? 136
5.1.3 What are the Indirect and Induced Effects of
Clean Energy Initiatives? 137
5.2 How Can States Estimate the Macroeconomic
Benefits of Clean Energy Initiatives? ....138
5.2.1 Step 1: Determine the Method of Analysis and
Level of Effort 138
5.2.2 Step 2: Quantify Expenditures and Savings
from the Clean Energy Initiative 148
5.2.3 Step 3: Apply the Method to Quantify
Macroeconomic Effects 153
5.3 Case Studies 154
5.3.1 New York: Analyzing Macroeconomic Benefits
of the Energy $mart Program 154
5.3.2 Illinois: Analyzing the Macroeconomic Benefits
of Clean Energy Development 156
Sampling of State Clean Energy Analyses by Type
of Analytic Method 157
Information Resources 159
References 159
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 133
-------
STATES QUANTIFYING THE ECONOMIC BENEFITS
OF CLEAN ENERGY POLICIES
Wisconsin's Focus on Energy Program advances cost effective
energy efficiency and renewable energy projects in the state
through information, training, energy audits, assistance and
financial incentives. Its efforts are designed to help Wisconsin
residents and businesses manage rising energy costs, promote
in-state economic development, protect the environment and
control the state's growing demand for electricity and natural
gas over the short and long term.
The Wisconsin Department of Administration conducted an
evaluation of the economic impacts of the Focus on Energy
Program from its inception in 2002 through 2026. The analysis
involved:
1. Documentation and extrapolation of the net direct effects
of the program, such as program-related spending, energy
cost savings and spending on new equipment;
2. Application of a regional economic model (in this case, the
REMI model); and
3. Analysis of the implications.
The results indicate that the Focus on Energy Program provides
net benefits to the State of Wisconsin. Specifically, the analysis
estimates that between 2002 and 2026, the Focus on Energy
Program is expected to:
create more than 60,000 job-years (see the text box Job
Years Versus Jobs);
generate sales for Wisconsin businesses of more than
eight billion dollars;
increase value added or gross state product by more than
five billion dollars; and
increase disposable income for residents by more than
four billion dollars.
Source: Wisconsin Department of Administration, 2007.
Build support for their clean energy initiatives
among state and local decision-makers; and
Identify opportunities where meeting today's
energy challenges can also serve as an economic
development strategy.
This chapter helps states understand the issues and
methods for assessing the economic benefits of clean
energy options so that they may conduct and manage
analyses, review cost and benefit estimates presented to
them, and make recommendations about the clean en-
ergy options the state should explore or the appropriate
evaluation approaches and tools to use.
Section 5.1 explains how clean energy initiatives create
direct, indirect, and induced macroeconomic effects
on the economy and can achieve benefits. Section 5.2
presents steps, methods ranging from rule-of-thumb
estimates to rigorous dynamic modeling, and issues
states can consider using to conduct an analysis of
the potential macroeconomic benefits of clean energy
programs. Section 5.3 describes a sampling of state
macroeconomic analyses as case studies.
5.1 HOW CLEAN ENERGY INITIATIVES
CREATE MACROECONOMIC BENEFITS
Clean energy initiatives can result in macroeconomic
benefits through direct, indirect, and induced economic
effects. As implied by these terms, some of the macro-
economic benefits of clean energy investments accrue
to those individuals, businesses, or institutions directly
involved in the investment, while other benefits arise
in related economic sectors and society as a whole via
indirect and induced "ripple" (or "multiplier") effects.
The design and scope of the clean energy initiative
typically determine the direct and indirect effects.
The structure and composition of the state's econ-
omy determine the resulting indirect and induced
effects.
The direct effects of policies or programs that affect
energy demand, such as those that stimulate invest-
ments in energy efficient equipment by the commercial
or residential sectors, will differ from the direct effects
of those that affect the supply of energy, such as renew-
able portfolio standards. The direct effects of these
demand and supply programs are key inputs to mac-
roeconomic analyses. The indirect and induced effects
are determined once the direct effects interact with
the overall state or regional economy. When exploring
the direct, indirect, and induced costs and benefits of
clean energy programs, it is useful to consider how the
initiative affects other state economic policy objectives,
such as distributional equity, and to ensure that it both
affects the segments of the economy that were initially
targeted and minimizes negative ramifications (e.g., a
resulting loss in jobs in another sector, which would
have distributional effects).
Direct, indirect, and induced effects are described in
greater detail below.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 134
-------
WHAT ARE DIRECT, INDIRECT, AND INDUCED EFFECTS?
Most approaches for quantifying local economic impacts
characterize economic impacts based on direct, indirect,
and induced effects. The same terms are used in computable
general equilibrium and hybrid macroeconomic models
Direct effects are changes in sales, income, or jobs
associated with the on-site or immediate effects created by
an expenditure or change in final demand; for example, the
employment and wages for workers who assemble wind
turbines at a manufacturing plant.
Indirect effects are changes in sales, income, or jobs in
upstream-linked sectors within the region. These effects result
from the changing input needs in directly affected sectors; for
example, increased employment and wages for workers who
supply materials to the turbine assemblers.
Induced effects are changes in sales, income, or jobs created
by changes in household, business, or government spending
patterns. These effects occur when the income generated from
the direct and indirect effects is re-spent in the local economy;
for example, increased employment and wages for workers at
the local grocery store because turbine assemblers use their
increased wages to buy groceries.
5.1.1 WHAT ARE THE DIRECT EFFECTS OF
DEMAND-SIDE INITIATIVES?
Clean energy initiatives that affect the demand side of
energy services typically change the energy consump-
tion patterns of business and residential consumers by
reducing the quantity of energy required for a given
level of production or service. Demand-side initiatives
generally aim to increase the use of cost-effective ener-
gy efficiency technologies (e.g., including more efficient
appliances and air conditioning systems, more efficient
lighting devices, more efficient design and construction
of new homes and businesses), and advance efficiency
improvements in motor systems and other industrial
processes. Demand-side initiatives can also directly
reduce energy consumption, such as through programs
encouraging changing the thermostat during the hours
a building is unoccupied or motion-detecting room
light switches.
The direct macroeconomic effects of demand-side
energy efficiency initiatives arise from the expenditures
for goods and services used to implement the initia-
tives as well as the energy and other cost savings gener-
ated by the initiatives. These costs and savings include:
Energy cost savings: dollars saved by businesses
and households resulting from reduced energy
costs (including electricity, natural gas, and oil cost
Demand-side initiatives usually change the end-use efficiency
of energy consumption.
Supply-side initiatives usually change the fuel/generation mix
of energy supply resources.
CLEAN ENERGY INITIATIVES EXPAND LOCAL RENEWABLE
ENERGY MARKETS AND REDUCE ENERGY COSTS
From 2001-2006, New Jersey's solar market experienced
strong growth and saved solar owners an estimated $1.1 million
annually in total electricity costs, spurred by the Customer
On-Site Renewable Energy Program (CORE), which provides
rebates for renewable technologies (NJ BPU, 2005).
savings), potentially reduced repair and mainte-
nance costs, deferred equipment replacement costs,
and increased property values resulting from the
new equipment.
Program administrative costs: dollars spent op-
erating the efficiency initiative, including labor,
materials, and paying incentives to participants. It
is important to determine how the costs of a pro-
gram will be funded, such as through a surcharge
on consumer electricity bills. If they are funded
through general government revenues, it is helpful
to consider the impact of diverting funds from
other projects.
Household and business expenditures: dollars spent
by businesses and households for purchasing and
installing more energy-efficient equipment. For
policies supported by a surcharge on electric bills,
the surcharge is a cost to be included.
Sector transfers: increased flow of dollars to
companies that design, manufacture, and install
energy-efficient equipment, and reduced flow of
dollars to other energy companiesincluding
electric utilitiesas demand for electricity and
less-efficient capital declines.
These direct costs and savings shift economic activ-
ity among participants. For example, they affect the
purchasing power of participating consumers, the
profitability of participating businesses, and the profit-
ability of conventional power generators. Together, the
shifts caused by demand-side initiatives affect income,
employment, and overall economic output by:
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 135
-------
CLEAN ENERGY INITIATIVES EXPAND LOCAL ENERGY
EFFICIENCY MARKETS AND ATTRACT BUSINESS INVESTMENT
For example, from 1999-2005, the number of energy service
companies operating in New York State increased from fewer
than 10 to over 180 companies, spurred by the New York
Energy Smart Program (NYSERDA, 2006b).
Decreasing residential energy costs, and thereby
increasing the disposable income available for non-
energy purchases.1
Increasing income, employment, and output by
reducing the outflow of resources that leave the
state when it imports electricity.2
Increasing income, employment, and output by
stimulating the production and sale of energy-
efficient equipment by existing businesses within
the state.
Increasing income, employment, and output by
decreasing the cost of doing business and improv-
ing competitiveness.
Increasing income, employment, and output by
expanding the in-state market for energy efficiency
and attracting new businesses and investment.
Decreasing revenue for utilities due to the re-
duction in energy sales, unless the state's utility
revenue structures allow for program cost recov-
ery or financial incentives for energy efficiency
programs.3
5.1.2 WHAT ARE THE DIRECT EFFECTS OF
SUPPLY-SIDE INITIATIVES?
Supply-side clean energy policies and programs change
the fuel/generation mix of energy resources or other-
wise alter the operational characteristics of the energy
supply system. Supply-side policy measures generally
1 An increase in disposable income may be reduced by any program costs
imposed upon them. Generally, however, the net effect to, for example, con-
sumers of energy efficiency programs, is positive.
2 The magnitude of this impact can be especially significant in states that
import large fractions of their energy
3 California, Massachusetts, Minnesota, New York, and Oregon have offered
utilities the opportunity to benefit financially from operating effective energy ef-
ficiency programs. These financial incentives reward utilities based on the level
of energy savings produced and/or cost effectiveness of their energy efficiency
programs (SWEEP 2002). It is important to consider each individual state's
utility revenue structure when exploring the effect of clean energy programs.
JOB YEARS VERSUS JOBS
Studies present employment estimates in terms of jobs and
job years, and it is important to understand the difference. For
example, a study may predict the creation of 15 job years. This
is not the same thing as saying 15 jobs. Fifteen job years can
mean one job that lasts for 15 years or it can mean 15 jobs that
last for one year. It is important to explain carefully or question
what the study is showing for potential job impacts.
In addition, sometimes job results are presented as "net jobs"
or even simply "jobs." If an analysis of a clean energy program
refers to "net jobs," it means the study factored in any job
losses that may have occurred in non-clean energy related
sectors due to the policy (e.g., decrease in demand for coal)
and presents the impacts on jobs after those losses have been
subtracted from any increase. If the results are presented as
"jobs," clarification may be needed to determine whether the
jobs are gross or net jobs.
support the development of utility-scale renewable
energy (RE) and combined heat and power (CHP) ap-
plications, and/or clean distributed generation (DG).
The direct effects of supply-side initiatives arise from
the costs of manufacturing, installing, and operating
the RE or CHP equipment supported by the initiative,
as well as the energy savings and possible reduced
energy supply costs from fuel substitution among enti-
ties participating in the supply-side program and their
customers. The direct costs and savings of RE/CHP/
DG initiatives include:
Displacement savings: dollars saved by utilities from
the displacement of traditional generation, includ-
ing reduced purchases (either local or imports) of
fossil fuels and decreased operation and mainte-
nance costs from existing generation resources.
Waste heat savings: dollars saved by utilities or
other commercial/industrial businesses using
waste heat in CHP applications for both heating
and cooling purposes.
Program administrative costs: dollars spent operat-
ing the initiative, including labor, materials, and
paying incentives to participants. As with demand-
side initiatives, it is important to determine how
the costs of a program will be funded, such as
through a surcharge on consumer electricity bills.
Construction costs: dollars spent to purchase the
RE/CHP/DG equipment, installation costs, costs
of grid connection, and on-site infrastructure con-
struction costs such as buildings or roads.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 136
-------
WHAT ABOUT OTHER ECONOMIC BENEFITS,
SUCH AS AVOIDED CAPACITY INVESTMENT
AND LOWER PRICE VOLATILITY?
Clean energy initiatives, whether on the demand side or supply
side of the energy system, can create other direct economic
benefits to individual energy producers and society as a
whole. These benefitswhich are economic in character but
arise specifically in the energy sectorinclude increased fuel
diversity, transmission reliability, avoided future investment in
fossil-fuel generating capacity, reduced wholesale electricity
price volatility, reduced fossil-fuel prices, and reduced
transmission congestion and losses.
Assessing these benefits requires different methods from those
used to assess the benefit mechanisms described in this chapter.
These benefits and their assessment are covered in Chapter 3.0,
Assessing the Electric System Benefits of Clean Energy.
Operating costs: dollars spent to operate and main-
tain the equipment during its operating lifetime
and the cost of production surcharges applied to
consumers.
The expenditures and savings associated with supply-
side clean energy initiatives shift economic activity
among purchases of fuels, business activity in RE/
CHP/DG generation, and business activity in existing
generation. Together, the shifts caused by supply-side
initiatives increase income, employment, and eco-
nomic output in the state through the:
Construction and operation of new clean energy-
based power facilities.
Stimulation of economic activity in the state's exist-
ing renewable energy industry for both in-state
and export markets.
Expansion of the in-state market for renewable
energy services and attraction of new businesses
and investment.4
Reduced outflow of dollars for fossil-fuel imports
(or increased inflow of dollars for fossil-fuel ex-
ports if state is a net fossil-fuel exporter), enabling
those dollars to remain within the state.
Increased application of CHP, in particular, by
reducing the cost of doing business and improving
overall competitiveness for non-energy companies.
4 See also, MTC (2005) and Heavner and Del Chiaro (2003) for additional
information on evaluating EE/RE market potential and fostering so-called
"clean energy clusters."
WHY QUANTIFY INDIRECT AND INDUCED EFFECTS?
Quantifying the full rangedirect, indirect, and inducedof
the macroeconomic benefits from clean energy initiatives
will maximize the potential value of the policy analysis. For
example, the University of Illinois' analysis in 2005 of the
proposed Illinois Sustainable Energy Plan estimated that the
direct outlays and savings for the plan would provide the
following benefits to the state of Illinois by 2020:
A $7 billion net increase in economic output,
A $1.5 billion net increase in personal income, and
43,000 net new jobs.
While these benefits are certainly substantial, the study further
estimated the following combined direct and indirect benefits
by 2020:
An $18 billion net increase in economic output,
A $5.5 billion net increase in personal income, and
191,000 net new jobs
In this case, the more robust quantification of macroeconomic
benefits, as opposed to simply quantifying direct benefits,
led to a substantially different appreciation of the economic
significance of the program to the State of Illinois. (Bournakis
and Hewings et al., 2005.)
5.1.3 WHAT ARE THE INDIRECT AND
INDUCED EFFECTS OF CLEAN ENERGY
INITIATIVES?
The distinction between demand-side initiatives and
supply-side initiatives is a key factor in understanding
the direct effects of clean energy initiatives, but this
distinction is not necessary to describe indirect and in-
duced effects. The indirect and induced effects of clean
energy initiatives arise, respectively, from changes in
sectors that are economically linked to the directly
affected sectors and from changes in the purchases
of retail goods and services by the employees of the
businesses in which the direct and indirect economic
effects occur.
Indirect Effects
Indirect effects result from "upstream" changes in busi-
ness activity among firms supplying goods and services
to industries directly involved in the clean energy
initiative. For example, the construction of roads and
foundations for a wind farm requires purchases of
asphalt and cement from other economic sectors. Each
of those other industries must also make purchases to
support its own operations, and so forth.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 137
-------
There also can be "downstream" indirect effects, as the
regional economy responds to lower energy costs, a
more dependable energy supply, and a better economic
environment fostering expansion and attracting new
business growth opportunities.
In a state-level macroeconomic impact analysis, the
fraction of all of the inter-industry purchases that
occur within the state comprises the indirect effects.
These purchases, in turn, affect income, employment,
and economic output in those intermediate sectors.
The ability of the states economy to provide the goods
and services needed to implement the initiative is a key
factor affecting the quantity of in-state indirect effects.
In general, a larger, more diverse economy will keep a
greater share of the indirect purchases within the state
(i.e., the indirect multiplier effects would be larger).
For example, a study of the economic benefits of clean
energy in New England for the Regulatory Assistance
Project noted that "if there were a substantial indig-
enous renewable generator manufacturing and main-
tenance industry in New England, then the projected
impacts would be larger" (RAP, 2005).
Induced Effects
Induced effects result from the additional purchases
of goods and services by households and governments
that are affected directly and indirectly by the clean
energy initiative as described above (e.g., increased
wage income generated from direct and indirect effects
is re-spent by individuals; taxes generated by direct and
indirect effects are re-deployed by governments). These
outlays, in turn, lead to changes in income, employ-
ment, and economic output in all economic sectors.
5.2 HOW CAN STATES ESTIMATE
THE MACROECONOMIC BENEFITS OF
CLEAN ENERGY INITIATIVES?
Assessing the state-level macroeconomic benefits of
clean energy initiatives involves measuring changes in
the flow of dollars to households and businesses at the
state level. Changes in these flows can be estimated as
gross impacts (changes without adjustment for what
would have occurred anyway) or net impacts (changes
over and above what would have occurred anyway).
The macroeconomic impacts of clean energy initiatives
can also be evaluated for cost-effectiveness. Cost-
effectiveness refers to the benefits generated per dollar
of program costs.
Quantifying the macroeconomic effectswhether on
a gross, net, or cost-effective basisprovides an aggre-
gate measure of the magnitude of the benefits achieved
by the initiative. A state can follow several basic steps
to analyze the macroeconomic benefits of clean energy
initiatives:
1. Determine the method of analysis, the desired
level of rigor, and the desired level of detail about
geographic and industrial sectors.
2. Quantify the direct costs and savings associated
with the initiative.
3. Apply the previously determined method to quan-
tify the macroeconomic impacts created by those
costs and savings.
Each of these steps is discussed in more detail below.
5.2.1 STEP 1: DETERMINE THE METHOD OF
ANALYSIS AND LEVEL OF EFFORT
Several methods are available to states for quantify-
ing the macroeconomic effects of their clean energy
initiatives. They range in complexity from using basic
approaches or tools for screening purposes to sophis-
ticated modeling tools for more rigorous dynamic
modeling approaches. All of these methods involve
predictions, inherent uncertainties, and numerous as-
sumptions. In selecting the most appropriate method,
states can consider many different factors, including
time constraints, cost, data requirements, internal staff
expertise, and overall flexibility and applicability. For
example, a state looking to quickly compare many
policy options to get an approximate sense of their
costs or benefits as part of a stakeholder process would
select a different tool than a state tasked by its governor
or legislature to determine the sector-specific impacts
of a particular policy or strategy. The latter situation
would likely require a more rigorous analysis.
Consequently, it is useful for state policy makers to
understand the basic differences between the different
models and approaches, their strengths and weakness,
and their underlying assumptions. The following sec-
tions introduce the basic concepts associated with wide-
ly accepted screening tools and more advanced models
for macroeconomic analysis of clean energy initiatives.
Table 5.2.1 describes the advantages and disadvantages
of each method and when it is appropriate to use.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 138
-------
TABLE 5.2.1 COMPARISON OF BASIC AND SOPHISTICATED APPROACHES FOR QUANTIFYING
MACROECONOMIC EFFECTS OF CLEAN ENERGY INITIATIVES
Sample Tools or
Type of Method Resources Advantages Disadvantages When to Use this Method
Basic Approaches:
Rule-of-thumb
Rule-of-thumb
Factors
May be transparent Overly simplified -When time and
. Requires minimal input assumptions
estimates and Job and Economic data, time, technical Approximate results
. Screening models Development Impact expertise, and labor.
Sophisticated
Approaches:
Input-Output;
Econometric;
Computable
General
Equilibrium; and
Hybrid Models
(JEDI) Model
RMI Community
Energy Opportunity
Finder
Renewable Energy
Policy Project Labor
Calculator
. IMPLAN,
. RIMS II
RAND econometric
model
BEAR
REMI Policy Insight
Inexpensive, often free.
More robust than basic
May be inflexible.
resources are short
For high-level.
preliminary, analyses
To get quick estimates of
employment, output and
price changes
When screening a
large number of policy
options to develop a
short list of options for
further analysis.
May be less transparent When policy options are
modeling methods. than spreadsheet well defined
May be perceived as
more credible than
basic methods.
Provides detailed
results
May model impacts
over a long period of
time
May account for
dynamic interactions
methods.
May require
extensive input
data, time, technical
expertise, and labor
commitments.
Often high software
licensing costs.
Requires detailed
assumptions that can
within the state/ significantly influence
regional economy.
results.
When a high degree of
precision and analytic
rigor is desired
When sufficient data,
time and financial
resources are available.
Basic Approaches for
Macroeconomic Impact Analysis
At the simpler, less resource-intensive level, screening
tools and approaches provide quick, low-cost analyses
of policies and require less precise data than needed
for a rigorous, advanced analysis. These screening
methods provide rough estimates of impacts and give
a sense of the direction (i.e., positive or negative) and
magnitude of the impacts upon the economy. They
provide a useful screening device when many options
are under consideration and limited resources are
available to conduct advanced analyses. For example, a
state considering a lengthy list of climate change miti-
gation options can use a screening tool to help rank the
candidates to create a short list of options that warrant
further analyses with more sophisticated tools. Screen-
ing approaches, such as rule-of-thumb job factors and
tools (e.g., NREL's JEDI model, the RMI Community
Energy Opportunity Finder, and REPP's Labor Calcu-
lator), are described below.
Rule-of-Thumb Economic Factors
States can apply rules of thumb or generic economic
factors to their program results to estimate the eco-
nomic impacts of clean energy measures in their states.
These rules of thumb are typically drawn from more
rigorous analyses and can be used when time and re-
sources are limited. However, they provide only rough
approximations of clean energy program impacts and
so are most applicable for use as screening-level tools
for developing preliminary benefit estimates and for
prioritizing potential clean energy activities. Table 5.2.2
lists several rules of thumb that states have used to
estimate the income, output, and employment impacts
of energy efficiency and renewable energy programs.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 139
-------
TABLE 5.2.2 RULES OF THUMB FOR ESTIMATING INCOME, OUTPUT, AND EMPLOYMENT IMPACTS
OF CLEAN ENERGY ACTIVITIES
Rule of Thumb
TYPE OF IMPAr
^me/Output
1 MW of wind generated requires $1 billion
investment in wind generator components.
REPP, 2005
http://www.repp.org/articles/static/l/binaries/Ohio_Manufacturing_Report_2.pdf
$1 spent on concentrated solar power in California
produces $1.40 of additional GSP.
Stoddardetal., 2006
http://www.nrel.gov/docs/fy06osti/39291.pdf
$1 spent on energy efficiency in Iowa produces
$1.50 of additional disposable income.
Weisbrodetal., 1995
http://www.edrgroup.com/library/energy-environment/iowa-energy.html
$1 million in energy savings in Oregon produces $1.5
million of additional output.
Grover, 2005
http://www.oregon.gov/ENERGY/CONS/docs/EcoNW_Study.pdf
TYPE OF IMPACT: Employment
$1 million in energy savings in Oregon produces
about $400,000 in additional wages per year.
Grover, 2005
http://www.oregon.gov/ENERGY/CONS/docs/EcoNW_Study.pdf
$1 billion investment in wind generator components
creates 3,000 full-time equivalent (FTE) jobs.
REPP, 2005
http://www.repp.org/articles/static/l/binaries/Ohio_Manufacturing_Report_2.pdf
$1 million invested in energy efficiency in Iowa
produces 25 job-years.
Weisbrodetal., 1995
http://www.edrgroup.com/library/energy-environment/iowa-energy.html
$1 million invested in wind in Iowa produces 2.5 job-
years.
Weisbrodetal., 1995
http://www.edrgroup.com/library/energy-environment/iowa-energy.html
$1 million invested in wind or PV produces 5.7 job-
years vs. 3.9 job-years for coal power.
Singh and Fehrs, 2001
http://www.repp.org/articles/static/l/binaries/LABOR_FINAL_REV.pdf
1 GWh of electricity saved through energy efficiency
programs in New York yields 1.5 sustained jobs.
NYSERDA, 2008
http://www.nyserda.org/pdfs/Combined Report.pdf
$1 million of energy efficiency net benefits in
Georgia produces 1.6-2.8 jobs.
Jensen and Lounsbury, 2005
http://www.gefa.org/Modules/ShowDocument.aspx?documentid=46
As shown in Table 5.2.2, for example, the Renewable
Energy Policy Project (REPP) estimates that every $1
billion of investment in the components that make
up wind generators creates 3,000 full-time equivalent
(FTE) jobs. REPP also finds that every megawatt (MW)
of wind requires a $1 billion investment in the genera-
tor components (REPP, 2005). If a state has estimated
the amount of renewable (wind) electricity that will
be generated from its clean energy programs, it can
use these factors to determine the amount of jobs that
could be created.
The New York State Energy Research and Develop-
ment Authority (NYSERDA) has developed a similar
jobs factor for energy efficiency programs. It estimates
that every GWh of electricity saved through energy
efficiency programs yields 1.5 sustained jobs.5 This fac-
tor is derived from a more sophisticated analysis of the
macroeconomic impacts of the New York Energy Smart
Program through 2007. This analysis estimated that the
program had created, on average, 4,700 net jobs each
year between 1999 and 2007 while saving about 3,164
GWhs in electricity (NYSERDA, 2008). Dividing the
5 By sustained, it means that the job is expected to last 15 years.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 140
-------
number of jobs by the number of GWhs saved through
energy efficiency measures yields an average number of
net sustained jobs, about 1.5, for each GWh saved. New
York uses this number to generate rough estimates of
the job impacts of new or expanded energy efficiency-
related programs under consideration. For example,
when New York announced its 15 by 15 initiative, which
set a goal of reducing energy demand by 15 percent or
27,300 GWhs through energy efficiency, NYSERDAs
rule of thumb was used to estimate that the initiative
was expected to create about 41,000 jobs in the state.
These rule-of-thumb factors can be handy when time
and resources for more rigorous analysis are limited.
As shown in Table 5.2.2, however, the range of values
is wide. For this reason, it is very important to under-
stand any biases that may be inherent in the rule of
thumb before using them. For example, factors can be
based on outdated information and would be affected
by changes in construction and material costs that have
occurred since the factor was derived. Alternatively,
factors may not take into consideration that the funds
are likely to have come from elsewhere in the economy
and may result in negative impacts. For example, the
REPP wind-related factor described above may not
consider that the $1 billion investment could have been
taken from another sector in the state or the United
States as a whole, which may now experience job losses.
There is an opportunity costthe value of the next best
alternative forgonethat states should consider when
taking resources from one place in the economy and
investing them in something different, in this case clean
energy. In addition, it is not clear if the 3,000 jobs are
net or gross. That is to say, it is not apparent whether
the numbers reflect job losses that may occur in other
sectors. It also is not obvious whether any additional
price increases that the consumer would have to pay for
renewable energy have been reflected in the analysis.
For energy efficiency programs, there are similar ques-
tions to consider when using a factor. When a state
implements a program for energy efficiency through
surcharges to rate payers, it is taking money away
from the consumers that it would have spent on other
goods, possibly creating job losses, and investing them
into the energy efficiency program, possibly creating
job increases.
Key questions to consider when using a rule-of-thumb
estimate include:
How recent are the construction and material costs
used in the factor?
USING JEDI: THE CASE OF WIND POWER IN
UTAH COUNTY, UTAH
Wind power has been proposed in Utah as a way to diversify the
state's electricity generation. Utah State University used JEDI
to inform decision makers about the likely impact of five wind
capacity scenarios: 5 MW, 10 MW, 14.7 MW, 20 MW, and 25 MW.
Economic and demographic information was obtained from
three sources: (1) the Economic Development Corporation of
Utah (EDCU); (2) IMPLAN multipliers for Utah county supplied by
NREL; and (3) two local wind developers. These data allowed the
study to dictate cost and other inputs specific to their scenarios.
The results of the JEDI analysis indicated promising economic
opportunities for wind power in Utah. For example, the
proposed Spanish Fork project (14.7 MW) would produce 46
total new jobs, $1.2 million in wage earnings, and $4.2 million in
economic output during the construction phase of the project
(Mongha etal., 2006).
Does it include the opportunity costs (lost jobs,
reduced earnings, spending or GSP) that occur
because the money for the clean energy program
was taken from elsewhere in the economy?
If the rule of thumb is related to employment, is
the estimate it generates given in jobs or job years
(for more information, see text box Job Years Ver-
sus Jobs earlier in this chapter).
Does the rule of thumb reflect any price increases
consumers may have to pay for the technology or
program?
Typically, these are the types of issues addressed in
more rigorous analysis but it is important to be aware
of any limitations associated with rule-of-thumb
factors. Because of these oversimplifications, rule-of-
thumb factors are best recognized as screening-level
tools that can provide preliminary estimates.
Screening Tools
Job and Economic Development Impact (JEDI)
Model for Wind Projects
The U.S. Department of Energy/National Renew-
able Energy Laboratory (DOE/NREL) developed a
spreadsheet-based model, JEDI, for estimating the local
economic effects of the construction and operation of
wind power plants. JEDI is designed to be user-friendly
and does not require experience with spreadsheets or
economic modeling. The model was originally devel-
oped with state-level parameters, but it can also be
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 141
-------
THE IMPORTANCE OF ACCURATE ENERGY DATA
Accurate and complete state energy data are often missing
or incomplete, but are a crucial input to any multiple benefit
analysis. States do not always have dynamic energy sector
representation and must rely on spreadsheet-level analysis.
used for county and regional analyses. Users enter ba-
sic information about the wind plant project (e.g., the
project's state, county, or region; the year of construc-
tion; the size of the facility), and JEDI calculates the
project cost as well as the jobs, income, and economic
output that will accrue to the state, county, or region
being analyzed. The project cost calculations are based
on default expenditure patterns derived from numer-
ous wind resource studies. The user can replace these
default values with project-specific information, such
as costs and expenditures, financing, taxes, and local
share of spending (Goldberg et al., 2004).
JEDI uses input-output analysis to evaluate the direct,
indirect, and induced macroeconomic effects from the
project expenditures. This type of analysis quantifies
relationships among industries in a state, regional, or
national economyi.e., showing how sales of goods
and services in one industry lead to purchases or sales
of goods and services in other industries. These rela-
tionships are depicted as state-specific multipliers that
show how the effects of an investment multiply beyond
the original transaction. The multipliers are adapted
from year 2000 data used in the IMPLAN* Professional
model, an input-output modeling tool described below
in Sophisticated Modeling Methods for Macroeconomic
Impact Analysis.
JEDI outputs should not be considered precise values,
but rather an indication of the magnitude of potential
economic development impacts. Structural characteris-
tics that limit the accuracy of JEDI's results include the
following:
JEDI outputs are presented as aggregate impacts
without sector specificity.
JEDI is a static model and cannot account for fu-
ture changes in wind power plant costs, changes in
industry, or personal consumption patterns in the
economy.
Analyses are specific to wind power plants and
therefore represent a gross analysis that does not
reflect net impacts associated with alternative uses
of the expenditures.
Analyses do not account for changes in electricity
prices or end-user electricity bills that could result
from developing the wind power plant.
Analyses assume that plant output generates suffi-
cient revenues to accommodate the equity and debt
repayment and annual operating expenditures.
JEDI does not calculate "net jobs" or otherwise
reflect the opportunity cost of alternative uses of
investment.
http://www.nrel.gov/' apply ing_technologies/
market_economic_mt.html
RMI Community Energy Opportunity Finder
The Rocky Mountain Institute (RMI) Community En-
ergy Opportunity Finder is an interactive website tool
that provides a preliminary analysis of the potential
benefits of implementing energy efficiency or renew-
able energy in a particular community. This tool has
the following characteristics:
Is designed to perform an initial evaluation of the
opportunities for energy efficiency and renewable
energy projects in the community.
Guides the user through the process of collecting
energy use data for the local community and then
calculates potential energy savings, dollar savings,
and job creation that could be achieved through
the energy efficiency or renewable energy project.
Includes many calculations and assumptions based
on published literature and substantial experience
from dozens of energy experts.
Can produce a reasonable estimated range of ben-
efits from a small core of energy use data.
Is limited by using largely default values and other
information not necessarily specific to the project
being analyzed.
Finder is intended to provide an overall sense of the
potential benefits of energy efficiency and renewable
energy options in a community, but should not serve
in place of a detailed audit of each area or building
where energy is used. A variety of cities, utilities, and
education programs have used Finder as a screening
tool. Examples of Finder applications are available at
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 142
-------
USING REPP LABOR CALCULATOR: THE CASE OF
NEVADA'S RPS
As part of its 1997 restructuring legislation, the Nevada
legislature established an RPS that included a 5% renewable
energy requirement in 2003 and a 15% requirement by 2013. The
Nevada American Federation of Labor-Congress of Industrial
Organizations (AFL-CIO) used the REPP Labor Calculator to
estimate the job diversification effects of the RPS (IREC, 2005).
To use the calculator, AFL-CIO had to make a number of
assumptions, including assumptions to estimate electricity
sales by technology type, which were then used to estimate the
installed capacity of each renewable technology.
The results of their analysis showed that, from 2003-2013, the
RPS would create 27,229 total, direct full-time-equivalent (FTE)
jobs. Of these jobs, 19,138 are estimated to be manufacturing
jobs while 8,092 are installation and O&M jobs. These are
direct jobs and do not account for any indirect or induced
employment effects (AFL-CIO, 2002).
the RMI website. RMI is currently working on revising
Finder and developing related web-based tools, http://
www. energyfinder.org/
REPP Labor Calculator
The Renewable Energy Policy Project (REPP) has
developed a tool that calculates the number of direct
jobs resulting from state programs, such as an RPS pro-
gram, that accelerate renewable energy development.
The Labor Calculator is based on a survey of current
industry practices related to manufacturing, instal-
lation, and operation and maintenance activities for
renewable technologies. The spreadsheet-based format
of the calculator provides a transparent framework that
lays out all of the labor data and program assumptions.
The user specifies the required installed capacity to
meet the renewable energy program requirements (e.g.,
an RPS), and the calculator determines the number
and type of jobs in each renewable activity area by year
per installed MW of capacity. The Labor Calculator
estimates the total direct labor required to manufac-
ture, install, operate, and service several types of clean
energy projects, including wind power, distributed
solar PV systems, biomass fuel production for use in
biomass co-fired coal plants, and geothermal power
plants. REPP is currently developing information to
expand the Labor Calculator to include other biomass,
geothermal, and solar thermal technologies.
MODELING ENERGY-ECONOMY INTERACTIONS:
BOTTOM-UP VS. TOP-DOWN
Bottom-Up and Top-Down analyses are the two primary
approaches for modeling energy-economy relationships. The
major differences between these approaches are the emphases
placed on a detailed technologically based representation of the
energy system, and the representation of the general economy.
Bottom-up models include a detailed representation of the
energy sector in the form of an energy technology matrix,
where each technology is represented by engineering cost
and performance characteristics. These models are capable
of capturing substitution among labor, capital, and fuel
inputs among technologies, and other structural changes in
the energy sector in response to a given stimulus or policy
constraint (Loschel, 2002). These models, however, generally
do not assess how energy system changes spill over to other
economic sectors and generate macroeconomic or general
equilibrium effects. Bottom-up models are also limited in
their ability to represent the influences of non-energy markets
on cost and performance dynamics of the energy system
technologies (Bohringer, 1998; Loschel, 2002).
Top-down models represent the energy sector in a more
aggregate way and account for how the energy sector interacts
with the rest of the economy. Rather than specifying energy
technologies according to their engineering characteristics,
top-down models usually represent technologies using
aggregate production functions that capture substitution
among technologies in response to price changes (i.e.,
substitution effects). In addition, top-down models usually
employ an input-output (I-O) table to simulate supply-demand
interactions and the reallocation of all goods and services
across the economy. All of the sophisticated modeling methods
described below are, fundamentally, top-down models.
The REPP tool is a job calculator, not an economic
model. It shows direct gross job effects that could be
captured by a state, but does not account for indirect or
induced secondary effects, http://www.repp.org/index.
html
Sophisticated Modeling Methods for
Macroeconomic Impact Analysis
The screening tools described above provide relatively
simple approximations of the economic feasibility and
impact of clean energy initiatives. They are often easy
to use, and results can be produced relatively quickly.
However, these tools do not typically provide a suf-
ficient level of sophistication to evaluate substantial in-
vestments in clean energy initiatives. Development and
implementation of clean energy initiatives at the state
level generally require a more comprehensive analysis
of the macroeconomic effects of alternative clean
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 143
-------
TABLE 5.2.3 OVERVIEW OF SOPHISTICATED MODELING APPROACHES AND TOOLS FOR
STATE ECONOMIC ANALYSIS
Example of
State Tools
Advantages
Disadvantages
Considerations
METHOD: Input-Output (also called multiplier analysis)
IMPLAN
Quantifies the total economic
effects of a change in the demand
for a given product or service.
Can be inexpensive.
Static; multipliers represent only
a snapshot of the economy at a
given point in time.
Generally assumes fixed prices.
Typically does not account for
substitution effects, supply
constraints, and changes in
competitiveness or other
demographic factors.
METHOD: Econometric Models
RAND
Usually dynamic, can estimate
and/or track changes in policy
impacts over time.
Coefficients are based on
historical data and relationships,
and statistical methods can be
used to assess model credibility.
Historical patterns may not be
best indicator or predictor of
future relationships.
Some econometric models do not
allow foresight.
Important to
understand if model
is myopic or has
foresight.
METHOD: Computable General Equilibrium (CGE) Models
BEAR
Account for substitution effects,
supply constraints, and price
adjustments.
Not widely available at state level.
Most CGE models available at
state level are static, although a
few are dynamic.
Important to examine
how the energy sector
is treated within any
specific CGE model.
METHOD: Hybrid
REMI
Policy
Insight
Most sophisticated, combining
aspects of all of the above.
Dynamic, can be used to analyze
both short- and long -term
impacts.
Can be used to model regional
interactions.
Flexibility of looking at 2-, 3-, or
4-digit NAICS sectors.
Can be expensive, especially if
there is a need to analyze impacts
on multiple sub-regions (e.g.,
counties within a state).
Can require a fair amount of
massaging inputs, especially with
energy sector inputs.
Important to examine
how energy sector is
treated.
May need to update
default data to account
for most recent energy
assumptions .
When to
Use
Provides rich sectoral Short-
detail (NAICS-based). term
Could be appropriate if analysis.
the need is to analyze
detailed impacts by
sector.
Short-
and
long-
term
analysis.
Long-
term
analysis.
Short-
and
long-
term
analysis.
energy initiatives. Several well established models have
been developed to quantify the nature and magnitude
of the macroeconomic effects of clean energy invest-
ments. These approaches include input-output models,
econometric models, computable general equilibrium
models, and hybrid models. Table 5.2.3 compares key
characteristics among these four model types.
Input-Output Models
Input-output (I-O) models, also known as multiplier
analysis models, are useful for quantifying macroeco-
nomic impacts because they estimate relationships
among industries in a state, regional, or national
economy. Policy impacts in I-O models are driven by
changes in demand for goods and services.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 144
-------
WHAT IS AN ECONOMIC MULTIPLIER ("RIPPLE EFFECT")?
An economic multiplier, usually expressed as a ratio, captures
how much additional economic activity is generated in each
regional industry from a single expenditure (or change in final
demand) in another industry.
In I-O models, multipliers estimate the size of sector-specific
indirect and induced effects, as well as the economy-wide
totals. Multipliers can be derived separately for employment,
income, and economic output, and are interpreted differently
depending on the form of the multiplier.
In California, for example, a study found that each $1 invested
in new solar generation would result in an additional $0.50
of economic activity in California (this represents an output
multiplier of 1.5). This study also found that 1MW of solar
capacity would produce an additional 40 job-years. (Cinnamon
and Beach etal., 2005)
At the core of any I-O model is an input-output table,
which describes the flow of goods and services from
producers to intermediate and final consumers. The
I-O table in the most commonly used I-O models in
the United States (e.g., IMPLAN, RIMS II) comes from
national and regional public data sources such as the
Bureau of Economic Analysis' (BEAs) national I-O
table and regional economic accounts.
The strength of I-O based models is their ability to
quantify the total economic effects of a change in the
demand for a given product or service. In this context,
"total" means the cumulative direct, indirect, and
induced effects. The I-O model produces a set of mul-
tipliers that describe changes in employment, output,
or income in one industry given a demand change in
another industry. It is important to note, however, that
the multipliers derived from I-O models only represent
a snapshot of the economy at a given point in time.
Due to their static nature, I-O models generally assume
fixed prices and do not account for substitution effects
and changes in competitiveness or other demographic
factors; thus they are suitable for static or short-term
analysis only (RAP, 2005).
In an analysis of the impacts of the Oregon Energy
Tax Credits, the modelers determined that the I-O ap-
proach was most appropriate for a short-term analysis.
With the IMPLAN model, they estimated that the net
impacts of the tax credits in Oregon for the year 2006
were an increase in:
Gross state product of more than $142 million.
Jobs by 1,240.
Tax revenue of nearly $10 million.
When it came to extrapolating the results into the
future, however, they acknowledged that "estimating
the long-term impacts taking into account regional
changes in energy efficiency and the subsequent impact
on economic output requires a much more extensive
dynamic modeling exercise (Grover, 2007)." Additional
studies that use input-output models are listed in the
resource section at the end of this chapter.
Econometric Models
Econometric models are a set of related equations that
use mathematical and statistical techniques to analyze
economic conditions both in the present and in the
future. Econometric models find relationships in the
macro-economy and use those relationships to forecast
how clean energy initiatives might affect income,
employment, output, and other factors. For example,
energy demand may be related to the price of fuel, the
number of households, and/or the weather but not
to individual income levels. These models examine
historical data to identify those relationships and make
predictions about the future.
Econometric models generally have an aggregate supply
component with fixed prices, and an aggregate demand
component. The models' regression coefficients are
similar to the multipliers produced by I-O models in
the sense that they describe how one component of the
economic system changes in response to a change in
some other component of the economic system. Most
econometric models use a combination of coefficients,
some of which are estimated from historical data, and
others that are coefficients obtained from other sources.
A key strength of econometric models is that they can
estimate and/or track changes in policy impacts over
time. Another strength is that consistency between the
econometric model structure (developed for analysis)
and the underlying economic theory can be evaluated
using statistical methods. For example, because histori-
cal data are used to generate specific coefficients that
reflect the observed relationships between variables,
statistical methods can be used to test whether the
observed historical data lend support to the (theoreti-
cally) hypothesized relationships between variables.
This requires the structure of an econometric model
to be formulated first based on economic theory and
then the model's coefficients estimated using historical
data, rather than developing the structure of the econo-
metric model itself based on the analysis of historical
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 145
-------
data (i.e., by developing the structure that best fits the
observed data).
Econometric models can be used for both long- and
short-term analyses. Because econometric models, in
general, rely heavily upon historical data as the pattern
for future behavior, the behavior projected is limited
because it neglects changes in consumer and business
conduct or investments that may occur when future
policies and price changes are anticipated. For example,
if a carbon standard were proposed today for imple-
mentation in five years, one might expect that firms
would begin making decisions about investments in en-
ergy sources and carbon-efficient technology that would
prepare them for when the mandatory provisions take
effect. A myopic econometric model might predict that
the actors will not alter their strategies until the manda-
tory provisions provide a "shock," even though they
would be able to anticipate the effect. Unless the econo-
metric model includes a mechanism for responding to
anticipated policy changes it may not be able to reflect
planning for implementation, thus missing investments
in new types of fuels or technologies or planning to
avoid last-minute capacity constraints and abandon-
ment of recently purchased equipment. The predicted
results of an unanticipated shock may be more negative
in the short term than something that is anticipated.
For this reason, users will need to be aware of the model
limitations and strongly consider choosing a tool with
foresight when conducting longer-term studies.
State-level econometric models are often developed by
universities, private consulting firms, or nonprofit or-
ganizations. For example, RAND Science and Technol-
ogy, a nonprofit institution, conducted an analysis for
the Commonwealth of Massachusetts to retrospectively
measure the economic benefit of energy efficiency im-
provements between 1977 and 1997. By looking at the
historical data with their econometric model, they con-
cluded that declines in energy intensity were associated
with increases in gross state product and that declines
in energy intensity can be an approximation of changes
in energy efficiency. They also concluded that govern-
ment investments in energy efficiency programs may
lead to improvements in gross state product. Through
statistical and mathematical equations, they could
explore the relationship between different key vari-
ables, such as energy intensity, gross state product, and
government investments, and determine which ones
were statistically linked (Bernstein, 2002a). The list at
the end of this chapter provides additional examples of
state-level clean energy project analyses that have used
econometric models.
Computable General Equilibrium Models
Computable general equilibrium (CGE) models use
economic data to trace the flow of goods and services
throughout an economy and solve for the levels of
supply, demand, and price that satisfy the equilibrium
constraints across a specified set of markets. Unlike
econometric models, CGE models use a framework
based on the tenets of microeconomic general equi-
librium theory: market clearance and no excess profit.
Market clearance refers to the notion that all economic
output is fully consumed and that all labor and capital
are fully employed. The no excess profit condition as-
sumes that in perfect competition, firms will continue
to enter any economic market until excess profits (i.e.,
profits exceeding a normal rate of return on capital) are
diminished to zero. A result of this is that prices will
equal the marginal cost of producing a product. When
the baseline equilibrium is perturbed, for example, by a
clean energy tax incentive, a new market equilibrium is
created. Firms will enter and exit existing markets, and
the economy will move to a new equilibrium, including
adjusting prices and output throughout the economy.
In this way, CGE models can be useful for assessing the
economy-wide impacts of a clean energy policy.
Many CGE models are calibrated using data from a
Social Accounting Matrix (SAM). A SAM is an exten-
sion of an I-O table, including additional information
such as the distribution of income and the structure of
production. Unlike I-O models, CGE models are able
to account for substitution effects, supply constraints,
and price adjustments in the economy snapshot. That
is, CGE models do not necessarily use fixed coefficients
and fixed prices to determine the relationships between
a sector and its upstream and downstream sectors. Like
I-O models, most CGE models are static, although
some are dynamic.
CGE models are best used for long-term analyses
because they may not accurately depict the economic
impacts a state experiences on its way to the new equi-
librium. The CGE analysis estimates what the economy
will resemble in the new steady state. Particularly
when compared with a static CGE model, econometric
models are typically better at capturing those interim
economic changes that will occur between the policy
shock and the new equilibrium.
It is important to examine how the energy sector is
treated within any specific CGE model. While it may al-
low for substitution effects, it may not include an option
for consumers or firms to switch to renewable energy or
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 146
-------
Analyzing Conservation
Policies in Connecticut
In 2004, Connecticut analyzed the eco-
nomic impact of oil and natural gas con-
servation policies in Connecticut. The
state wanted to explore the impacts of fully
funding a program between 2005 and 2020
to increase the efficiency of oil and natural
gas for residential, commercial, and indus-
trial users.
Connecticut used a hybrid model, the REMI
Policy Insight model, for their analysis.
REMI is a frequently used proprietary model
in the US for analyzing state level policy ini-
tiatives. Because the model does not have
a detailed energy sector module to fully
capture the fuel-switching that would oc-
cur within the electricity sector, Connecti-
cut used outputs from an energy analysis
using an electricity dispatch modelICF
International's IPMto estimate the energy
changes used as inputs to Policy Insight.
The direct costs included cost increases
resulting from a 3% natural gas-use and oil-
use surcharge on residential, commercial,
and industrial users to pay for the program;
the savings to residential, commercial, and
industrial users due to reduced consump-
tion of natural gas and oil; the consumption
reallocation of other consumer goods due
to an increase in personal income; the loss
in sales to natural gas and oil firms due to
ECONOMIC GROWTH DUE TO CONSERVATION POLICIES IN
CONNECTICUT (CUMULATIVE 2005-2020)
Employment (Average Annual Increase)*
Output (Mil '96$)
GSP(Mil'96$)
Population
Real Disposable Personal Income (Mil '96$)
State Revenues (Mil '01$)
Oil & Natural Gas
2,092
3,094.90
2,033.01
3,604
1,749.42
382.13
Oil
430
82.80
266.21
717
294.81
66.75
Natural Gas
1,668
3,020.64
1,773.82
2,894
1,459.35
314.97
* Employment is the average annual increase from the baseline. Employment is not
cumulative and is based on output growth. Source: REMI, 2004.
reduced consumption; and the investment
in new equipment, construction, research,
and other sectors.
These direct effects were used as inputs to
the REMI model to determine the indirect,
induced, and overall effects of the program.
The model was able to break down the
results to determine the contribution the
oil conservation efforts and the natural gas
conservation efforts made to the overall
economic impact. For example, as shown
in the above table, the overall result of the
analysis showed economic benefits to the
state. The natural gas conservation ef-
forts, however, contributed more than the
oil programs to the overall benefits of the
program. Because the model contains very
detailed sector-specific information, the
analysts were able to determine that "The
disproportionate ratio between the oil and
natural gas policies is due to the higher loss
in demand for petroleum than for natural
gas... the loss in demand of oil is almost 6
times higher than the loss in demand for
natural gas" (REMI, 2004).
energy efficiency as a way to meet energy demand. In-
dividual models will handle this differently depending
upon the details (e.g., number of sectors) of the model.
CGE models are more readily available at the national
level than at the state level, and most CGE models are
highly aggregated. Some states, however, have devel-
oped and/or used state-specific CGE models to analyze
the impacts of clean energy initiatives.6 In California,
for example, the University of California at Berkeley
developed a dynamic CGE model, the Berkeley Energy
and Resources (BEAR) model. In addition to the core
CGE model, it includes extensive detail about the
energy sector and also estimates greenhouse gas emis-
sions. This model has been used to assess the potential
6 RTI International developed a CGE model (the Applied Dynamic Analysis
of the Global Economy (ADAGE) Model) that can be used to explore dynam-
ic effects of many types of energy, environmental, and trade policies, including
climate change mitigation policies. For more information on CGE models and
their application for macroeconomic impact analysis, see Sue Wing (2004).
impacts of state-level greenhouse gas mitigation poli-
cies in California. A recent analysis concluded that
nearly 50 percent of California's 2020 goal of reducing
emissions levels to 1990 levels could be achieved using
just a handful of options under consideration, while
increasing gross state product by 2.4 percent and creat-
ing more than 20,000 jobs (Roland-Hoist, 2006).
Hybrid Models
Hybrid models incorporate aspects of two or more of
the modeling approaches described above, with most
models linking an I-O model to an econometric model.
Most hybrid models used for energy-related analyses
are described as regional economic-forecasting and
policy-analysis models. These models are the most
sophisticatedand expensiveof the four categories
of models.
These models include five analytic elements: (1) output,
(2) labor and capital demands, (3) population and
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 147
-------
labor supply, (4) wages, prices, and profits, and (5)
market shares. The integrated structure of these models
allows them to capture everything from economic
migration to changes in relative prices and the overall
competitiveness of businesses in the economy. These
models also include dynamic frameworks that support
forecasting of both what will happen in response to an
initiative and when it will happen.
Of the general approaches described in this section, the
hybrid modeling approach offers the most flexibility
and detail in tailoring an analysis to estimate the effect
of a specific clean energy initiative on a states economy.
A user can specify and forecast numerous different
model inputs, including: industry output, industry
demand, government, investment and/or consumer
spending, employment, factor productivity, labor sup-
ply, production costs, business taxes and credits, fuel
and/or labor costs, wages, housing and consumer prices,
and market shares. The results of the complex, dynamic
simulations produced by hybrid models can be distilled
into net impacts on key economic policy indicators,
such as employment, income, and overall economic
output. Hybrid models can be effective at estimating
both the long- and short-term impacts of policies.
As with other models, it is useful to examine how the
energy sector and technological change are treated
within a hybrid model. Many states have found that
detailed energy-related analyses require energy model-
ing to be done separately and used as inputs to a hybrid
model. This can be a limitation of some hybrid models.
In addition, these models can be very complex, time-
consuming and expensive to run, and require signifi-
cant input data.
Hybrid models used for policy analyses include REMI
Policy Insight* (see text box Analyzing Conserva-
tion Policies in Connecticut), those developed by the
Regional Economics Applications Laboratory (REAL,
developed at the University of Illinois), the Illinois
Regional Econometric Input-Output Model (ILREIM),
and the Georgia Economic Modeling System (GEMS,
developed at the University of Georgia). A list of ad-
ditional state-level analyses conducted using hybrid
models is provided at the end of this chapter.
Comparison of Models Commonly Used by States
to Analyze Clean Energy Initiatives
Table 5.2.4 summarizes key aspects of the four model
typesinput-output, econometric, CGE and hybrid
that have been frequently used for energy-related
The direct, indirect, and induced macroeconomic benefits
arise from the outlays, energy, and dollar savings generated by
clean energy initiatives.
It is important for states to understand these outlays and
savings because they are key inputs for quantifying changes in
employment, income, and output.
policy analyses. State analysts can consider this model
information in deciding upon an appropriate model
for analyzing the macroeconomic benefits of clean en-
ergy initiatives. No one model is perfect for any given
analysis case, and the analyst may often choose a given
model because it has been used previously for analyses
within a state and certain individuals within the state
analytic community are more familiar with run specifi-
cation and interpretation of model outputs.
5.2.2 STEP 2: QUANTIFY
EXPENDITURES AND SAVINGS
FROM THE CLEAN ENERGY INITIATIVE
The second step in analyzing macroeconomic effects is
to quantify the direct expenditures and savings from
implementing the clean energy initiative. The expen-
ditures and savings are the primary inputs to the sub-
sequent analysis of macroeconomic effects on income,
employment, and output. As described in Sections
5.1.1 and 5.1.2, the specific expenditures and savings
that states need to consider in this step are different for
demand-side and supply-side initiatives. But generally
speaking, these expenditures and savings include esti-
mates of energy savings associated with the initiative
and data on expenditures by participating entities and
the costs of administering the program.
Key Considerations for Quantifying
Expenditures and Savings
States have found it useful to design a strategy to quan-
tify initiative expenditures and savings based on (1) the
design and nature of the initiative, (2) the attributes of
the states economy, and (3) the expected behavior of
the initiative participants. Several factors contribute to
the challenge of developing such a strategy. The analyst
can consider the following factors when establishing
the necessary data to estimate expenditures and savings
(DOA, 2001):
Expected energy savings or costs (e.g., oil, natural
gas, electricity) to consumers over time. To perform
an economic impact analysis, it is often important
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 148
-------
TABLE 5.2.4 COMPARISON OF MODELS FOR ESTIMATING MACROECONOMIC BENEFITS
General Model Category
Example*
Input-Output
IMPLAN
Econometric
RAND
BEAR
Hybrid
REMI Policy Insight
Model Characteristics
I-O Component
CGE Component
Econometric Component
Open/Closed Economy
Dynamic Modeling Capability
State and County Level Modeling
Major Data Sources
Yes
No
No
Both
No
Yes
BEA, BLS, CBP, and
Census
Modified I-O
Varies
Varies
Varies
Yes
Certain Models
Varies
Social Accounting
Matrix
Yes
Limited
Yes
Certain Models
Varies
Varies
Yes
Yes
Yes
Open
Yes
Yes
BEA, BLS, CBP, EIA
and Census
^
Industry Characteristics
SIC/NAICS Classifications
Sector Aggregation Options
Yes
Yes
Varies
Yes
Varies
Yes
Yes
Yes
Other Features
Trade Flows
Substitution Effects
Price and Wage Determination
Feedbacks on Competitiveness
Migration, Demographic Changes
Yes
No
No
No
No
Certain Models
Varies
Yes
Yes
Varies
Most
Yes
Yes
Yes
Varies
Yes
Yes
Yes
Yes
Yes
Impacts Measured
Employment
Income
Output
Value Added
Proprietary
Overall Cost, Complexity, and Capability
Yes Yes
Yes Yes
Yes Yes
Yes Yes
Yes Some
Medium High
Yes
Yes
Yes
Yes
Some
High
Yes
Yes
Yes
Yes
Yes
High
* Models names are included for illustrative purposes only, and do not imply an endorsement by EPA.
to translate any energy savings into dollars. This
monetization can be accomplished by applying
projections of prices for different energy types
(e.g., coal, oil, gas, electricity) to the profile of ex-
pected energy savings. For example, a policy that is
funded by a surcharge on electricity bills imposes
a cost on consumers but the energy efficiency
investments will result in energy cost savings. Both
will affect the economy. For more information on
calculating energy savings, see Chapter 2, Assess-
ing the Potential Energy Impacts of Clean Energy
Initiatives.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 149
-------
Expected clean energy investment and realization
rates in the short and long terms. This factor is par-
ticularly important with regard to energy efficiency
initiatives. In assessing the expected change in
energy use from the proposed initiative, it is help-
ful to break down the most likely level of energy
savings realization by participant group and/or
equipment type. This "intention" information may
be collected via a survey of potential participants
or estimated using program analyses.7 For example,
a program might expect to achieve 30 percent
penetration of a new technology in the residential
sector in the short run, but 60 percent in the longer
term. Both short- and long-term realization rates
can significantly affect the overall magnitude and
time profile of program effects. Therefore, states
may find it useful to analyze the potential impacts
of a program under different realization scenarios
and then focus program efforts on achieving the
optimal level of adoption over time.
The proportion of investment from individual par-
ticipants versus program funding. The energy sav-
ings from a program are partially reduced by any
up-front outlays by program participants. It is im-
portant to account for participants' expected out-
lays because these outlays will affect the economic
performance of the total program (including
outlays and savings for participants). Participants'
expenditures (and expected downstream savings)
will also influence program participation. It is also
an important factor to account for the amount and
source of program funding. Program expenditures
can affect the state economy; however, the nature
and extent of those effects will depend on where
the program funds come from (e.g., a system
benefits charge applied to electricity bills) and the
distribution of funds across different economic
sectors. For example, a state might implement an
energy-efficient water heater rebate program that is
funded through a surcharge on all electricity bills.
A portion of the amount paid will be returned
to some consumers in the form of rebates. These
rebates will cover some of the purchase cost of
the new water heater. In this instance, the invest-
ment in the new water heaters is paid by program
participants directly and electricity consumers
through the surcharge.
7 As a corollary, in estimating the energy savings to be achieved by a
program, it is also important to account for, and net out, the baseline energy
savings that would have occurred without the program.
The amount of initiative-related activity expected to
occur locally. For any type of spending/sales that
originates within a state, part of the dollars will flow
to businesses located in the state and part will flow
to businesses outside of the state. Accounting for
where those dollars flow is important because, to
the extent that program-related flows replace flows
that would have otherwise left the state, there is
potential for in-state net economic gain. This effect
is known as "import substitution," and is measured
by factors called "regional purchase coefficients"
(RPC). All four of the economic models shown in
Table 5.2.4 use RPCs to account for this effect.8
1 The expected useful life of the clean energy invest-
ment. Any estimation of program expenditures and
savings requires information on the useful life of
the products or services provided by the program.
The costs and savings associated with program
investments can be amortized over the expected
useful life of the product or service. For example, a
state program might promote the purchase of ener-
gy-efficient appliances but these appliances do not
last indefinitely. It is important to consider life of
the products when calculating potential long-term
benefits of a program. If one expects the program
to continue beyond the useful life of the initial
investments, the analysis can also account for re-
newed investments when estimating the long-term
character of program expenditures and savings.
The expected persistence of energy savings over time.
Estimation of expenditures and savings requires
assumptions regarding the persistence of the en-
ergy savings over time. This may be, for example,
an assumed annual loss of energy savings attribut-
able to factors such as deterioration of equipment
performance, removal of equipment, business clo-
sures, or other factors relevant to the persistence of
demand- or supply-side energy saving effects. Note
that the useful life of a clean energy investment is a
key determinant in the ultimate long-term persis-
tence of savings, but that the persistence of savings
can also vary over time during the useful life.
The expected economic benefits associated with
energy system, environmental, or public health ben-
efits. Potential energy system benefits, such as fuel
8 RPCs can be estimated for specific products or services based on analysis of,
for example, the extent to which a state has a disproportionately large or small
base of 'manufacturers providing the relevant types of energy-saving equip-
ment (DOA 2001). Alternatively, many economic models contain default RPC
values.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 150
-------
cost savings and avoided capacity or transmission
and distribution costs to the electricity generators
and/or distributors, are economic benefits that can
be estimated and included in an economic analysis
in addition to the energy cost savings to consumers
above. (For more information about energy system
benefits, see Chapter 3, Assessing the Electric System
Benefits of Clean Energy.) Likewise, environmental
benefits such as reductions in criteria air pollutants
can reduce the costs of complying with air quality
standards and yield human health benefits such
as avoided deaths, illnesses, and hospitalizations,
and reductions in lost work days due to illnesses.
As described in Chapter 4, an economic value can
be estimated for many of these benefits, and they
can be included in the economic analysis to ensure
adequate representation of the overall benefits in
the analysis.
In addition to these considerations, the appropriate
method and data for quantifying costs and savings are
influenced by the macroeconomic analysis method
selected in Step 1 and its associated data requirements.
Methods for Quantifying Direct
Expenditures and Savings
A wide range of methods can be used to quantify the
direct expenditures and savings of a potential clean
energy initiative (that go beyond those covered in
this Resource), and states often develop a customized
approach based on their specific needs and resources.
For a prospective analysis of expenditures and savings,
most methods involve projections using some model-
ing capability. Models available for prospective analyses
range from relatively simple, spreadsheet-based models
like Excelergy (see California example below) to more
rigorous and data-intensive models such as the Long
Term Industrial Energy Forecasting model (LIEF)
and the Integrated Planning Model (IPM*), an electric
power sector model (see Georgia and SWEEP exam-
ples). If an initiative has already been implemented, the
modeling approach can be supplemented with actual
expenditure and/or savings data from the program.
In such instances, analysts can use already-collected
program data on expenditures and savings as inputs to
a retrospective analysis of macroeconomic effects, or as
inputs to a prospective analysis of future expenditures
and savings (or both, as is done in Massachusetts -
described below). Including these actual expenditures
and savings likely will require some type of "mapping"
to defined economic sectors (e.g., by NAICS or SIC)
before being entered into the models.
Examples of the methods that states have used to
quantify expenditures and savings in prior analyses
of clean energy initiatives are presented below. The
first three examples (California, Georgia, and the
Southwest) describe instances where the analysis was
prospective and used modeling techniques to estimate
the expenditures and savings of potential clean energy
investments. The last example (Massachusetts) shows
how a state might use actual clean energy initiative
expenditures and savings to (1) estimate macroeco-
nomic effects retrospectively and (2) project the future
expenditures and savings from an initiative [also see
Grover (2005), NYSERDA (2006b), and Sumi et al.
(2003) for examples of using actual initiative data on
expenditures and savings to estimate macroeconomic
effects retrospectively]. More information on these
studies can be found in the resource section at the end
of the chapter.
California Concentrated Solar Power
A study of concentrated solar power (CSP) in Califor-
nia evaluated the potential benefitsin terms of direct
and indirect effects on employment, earnings, and
GSPof the deployment of 2,100-4,000 MW of CSP
from 2008-2020 (Stoddard et al., 2006). The outlay and
savings data needed to quantify the direct and indirect
effects of the project on employment, earnings, and
GSP included the dollars spent by the project in Cali-
fornia on materials, equipment, and wages.
The California study used data from the Excelergy
Model, developed and maintained by NREL, to esti-
mate the expenditures and savings generated in the
CSP scenario. The data used by Excelergy to determine
the expenditures and savings included the size of the
plants to be built and the time periods for construc-
tion. Excelergy is an Excel spreadsheet-based model
for solar parabolic trough systems that models annual
plant performance and estimates capital and O&M
costs. The data produced by Excelergy served as the
input data for the macroeconomic analysis.
The study found that the "high CSP deployment"
scenario would result in $13 billion in investment, of
which an estimated $5.4 billion is estimated to be spent
in California. Using RIMS II, the study found that this
in-state investment would have a gross impact of $24
billion on California GSP.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 151
-------
Georgia Energy Efficiency Potential Study
To assess opportunities for energy efficiency invest-
ments in the state of Georgia, the Georgia Environ-
mental Facilities Authority undertook a prospective
analysis of the macroeconomic effects of varying levels
of investment in energy efficiency in Georgia from
2005-2015 (Jensen and Lounsbury 2005).
The expenditure and savings data required for the
Georgia study included the costs of energy efficiency
equipment, customer energy bill savings, and program
administration and incentive costs. To quantify these
inputs, the study used ICF International's Energy Effi-
ciency Potential Model (EEPM) to estimate the poten-
tial for energy efficiency improvements through pro-
gram and policy interventions, and the expenditures
and savings associated with realizing that potential.
The EEPM model provided a detailed view of which
sectors, subsectors, and end uses provide the great-
est opportunity for energy efficiency improvements
in Georgia's economy by using end-use forecast data
along with industry data on the costs, applicability, and
longevity of energy efficiency measures. Within the
model, the extent to which energy efficiency measures
are adopted over time depends on the costs of energy
efficiency measures relative to supply-side options
and the intensity of the projected policy interventions.
This relationship allowed the analysis to account for
the energy savings and expenditures associated with
efficiency investments and program administration, as
well as the cost and revenue reductions experienced by
utilities from reduced demand for electricity or gas.
Since Georgia-specific data for end-use forecasts
and utility avoided costs were not publicly available,
the study used regional data from various sources,
including the U.S. Department of Energy, the Energy
Information Administration, and IPM model projec-
tions. Results from EEPM and IPM were used as inputs
to the Georgia Economic Modeling System (GEMS),
developed by the University of Georgia, to estimate
macroeconomic development effects.
The results of the GEMS analysis demonstrated that
investments in energy efficiency in Georgia would gen-
erate economic benefits. Specifically, the study explored
three policy scenarios to capture the energy efficiency
potential identified for Georgia: a minimally aggressive
scenario, a moderately aggressive scenario, and a very
aggressive scenario. The study concluded that each sce-
nario would achieve long-term net economic benefits in
Georgia including the creation of 1,500 to 4,200 new jobs
and an increase in real disposable income of $48 million
to $157 million by 2015 (Jensen and Lounsbury, 2005).
Southwest "High Efficiency" Study
The Southwest Energy Efficiency Project (SWEEP, 2002)
analyzed the macroeconomic effects of investments in
energy efficiency from 2003-2020 in southwestern states
(including Arizona, Colorado, Nevada, New Mexico,
Utah, and Wyoming). A "high efficiency" scenario was
developed in the study by first establishing the expected
level of energy savings and expenditures that would
comprise this scenario.
In the residential and commercial sectors, SWEEP ana-
lyzed the energy savings and efficiency expenditures for
the "high efficiency" scenario using the DOE-2.2 model,
developed by James J. Hirsch & Associates in collabora-
tion with the Lawrence Berkeley National Laboratory,
accounting for specific building characteristics, energy
use practices, state-by-state saturation and usage rates
for end-uses, and other assumptions. This analysis
included data from SWEEP, ACEEE, and EIA, among
others.
In the industrial sector, SWEEP used the Long-Term
Industrial Energy Forecasting (LIEF) model, along with
U.S. Census and EIA data, to analyze the cost-effective
electricity savings for the "high efficiency" scenario
versus a base case scenario. LIEF is a model developed
by the Argonne National Laboratory that uses three
key factors to estimate the cost-effectiveness and adop-
tion of energy efficiency measures in the industrial
sector: (1) the assumed penetration rate, (2) the capital
recovery factor, and (3) projected electricity prices. The
LIEF model contains a number of cost assumptions for
energy savings, and also has a number of parameters
that the user can specify.
These analyses revealed, for example, that the "high effi-
ciency" scenario would reduce average annual electricity
demand growth from 2.6 percent in the base case to 0.7
percent, thereby reducing electricity consumption 33
percent by 2020 versus the base case. These and other
savings would accrue with a total investment of $9 bil-
lion from 2003-2020. The macroeconomic effects of
these expenditures and savings were then evaluated for
their direct, indirect, and induced effects using the IM-
PLAN input-output model. Among the findings of the
IMPLAN analysis were increased regional employment
of 58,400 jobs and increased regional personal income of
$1.34 billion per year by 2020 (SWEEP, 2002).
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 152
-------
TABLE 5.2.5 SUMMARY OF ECONOMIC IMPACTS OF 2002 MASSACHUSETTS ENERGY EFFICIENCY
PROGRAM ACTIVITIES
Electricity Bill Impacts
Energy Savings
Total Participant Annual Energy Savings
Average Life of Energy Efficiency Measures
Total Participant Lifetime Energy Savings
Total Program Costs
Average Cost for Conserved Energy
$21.5 million
14 years
$249 million
$138 million
4.0 C/kWh
Demand Savings
Total Participant Annual Demand Savings
Systems Impacts
Savings to All Customers Due to Lower Wholesale Energy Clearing Prices
Economic Impacts
Number of New Jobs Created in 2002
Disposable Income from Net Employment in 2002
$1.2 million
$19.4 million
2,093
$79 million
Source: Division of Energy Resources, 2004.
Massachusetts Annual Report on Energy Efficiency
The Massachusetts Division of Energy Resources
(DOER) produces an annual report analyzing the
impacts of ratepayer-based energy efficiency programs
in the state. The 2004 report is a retrospective analysis
of the macroeconomic effects of investments in energy
efficiency made in 2002 (DOER, 2004).
To perform the macroeconomic analysis, the DOER
first determined the expenditures and savings for
the 2002 investments. Program expenditures in
2002 included administration, marketing, program
implementation, program evaluation, performance
incentives paid to the distribution companies, and
direct participant costs (2002 investments totaled $138
million). Program administrators collect these data on
a continuous basis. Savings included direct participant
energy savings and electricity bill reductions, which
were estimated using a combination of data from Mas-
sachusetts distribution companies, including participa-
tion rates, average energy use per participant, and elec-
tricity rate impacts for each customer sector specific
to each electric distribution company service territory.
The detailed expenditure and savings data were then
further disaggregated into industry-specific measures
using Bill of Goods data developed by a contractor.
Using the expenditure and savings inputs, the DOER
modeled the macroeconomic effects of 2002 program
investments on employment, disposable income, and
GSP using the REMI Policy Insight model. In addition,
the DOER used those same expenditure and savings
data in combination with the Energy 2020 model to
project the lifetime energy savings of the 2002 program
activities. Using these projected savings from Energy
2020 as inputs, the DOER used the Policy Insight mod-
el to estimate the future economic benefits reflected in
Table 5.2.5.
5.2.3 STEP 3: APPLY THE METHOD TO
QUANTIFY MACROECONOMIC EFFECTS
Once the direct expenditures and savings of a clean
energy initiative have been quantified, the final step is to
assess the aggregate macroeconomic effects of the initia-
tive by applying the screening tool or modeling method
selected in Step 1. With regard to policy implementa-
tion, many states have found the rigorous modeling
methods outlined in Section 5.2.1 to be most effective
in generating support for clean energy actions when a
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 153
-------
robust assessment of the full range of effects (i.e., direct,
indirect, and induced) is required. The application of
economic impact models to measure the effects of en-
ergy efficiency and renewable energy policies is widely
used and accepted across the nation (Sumi et al., 2003).
Regardless of the method, the macroeconomic effects
of a clean energy initiative are usually quantified in
comparison to a projected baseline scenario of eco-
nomic activity. Constructing a base case scenario, or
updating a default base case that may be included in
the model, is generally the first step in the process of
applying the macroeconomic analysis method.
Comparing the effects of the initiative to a baseline
enables quantification of the overall net impacts of the
initiative because the base case reflects what would
have occurred in the initiatives absence. Typically, the
baseline scenario characterizes a business-as-usual
forecast of energy use patterns and economic growth
within the state assuming the funds for the initiative
are reallocated to other government programs or BAU
consumer spending levels. If states choose to pursue
one or more of these methods, the base case should be
developed according to specifications associated with
that particular method of analysis. This Resource does
not explicitly cover methods for economic base case
scenario development.
The remaining steps in applying the method depend on
the method chosen and the state's customized model-
ing scenarios for their slate of clean energy initiatives.
These attributes will, in turn, influence how the results
of the analysis should be interpreted for policy pur-
poses. The steps taken by Connecticut in the analysis
of their conservation program are described in the text
box Steps in a Macroeconomic Impact Analysis: Con-
necticut's Oil and Natural Gas Conservation Policies.
5.3 CASE STUDIES
5.3.1 NEW YORK: ANALYZING
MACROECONOMIC BENEFITS OF THE
ENERGY $MARTSM PROGRAM
Benefits Assessed
Net jobs and job years
Personal income
Total output
Gross state product
STEPS IN A MACROECONOMIC IMPACT ANALYSIS:
CONNECTICUT'S OIL AND NATURAL GAS
CONSERVATION POLICIES
EPA and the State of Connecticut analyzed the impacts of
Connecticut's proposed oil and natural gas conservation
policies as part of the state's Climate Change Action Plan (CT
GSC, 2004).
Step 1: Determine the method and level of effort
Connecticut was interested in a dynamic analysis of
both the economic and demographic impacts of these
conservation policies over a 15-year time horizon.
Connecticut contracted with Regional Economic Models,
Inc. (REMI Policy Insight model) to model the policies
because REMI's capabilities were consistent with its
objectives and modeling needs.
Step 2: Quantify outlays and savings from the initiative
The outlays and savings to be captured by the REMI Policy
Insight model included oil and gas cost increases for users
resulting from the surcharge on oil and natural gas; savings
to oil and gas users due to reduced consumption of oil and
natural gas; consumption reallocation of other consumer
goods due to an increase in personal income; loss in sales
to natural gas and oil firms due to reduced consumption;
and investment in new equipment, construction, research,
and other sectors.
Data for the analysis were provided by an IPM study
conducted for Connecticut, Northeast States for
Coordinated Air Use Management (NESCAUM),
Environment Northeast, Institute for Sustainable Energy,
CT Department of Public Utility Control, CT Department
of Environmental Protection, CT Clean Energy Fund, and
United Technology Corporation.
Step 3: Apply the method to quantify macroeconomic benefits
REMI developed a baseline forecast using a 53-sector
model for Connecticut, along with three alternative
conservation policy scenarios.
The total macroeconomic effects of the policy scenarios
were presented using the following indicators:
employment, output, GSP, real disposable income, state
revenues, and population changes.
The implementation of CT's proposed oil and natural gas
conservation policy is pending legislative action.
Clean Energy Program Description
The New York Energy Smart public benefits program,
created in 1998 and administered by the New York State
Energy Research and Development Authority, promotes
energy efficiency across the commercial, industrial, and
residential sectors; advances renewable energy; provides
energy services to low income residents of New York;
and conducts research and development (NYSERDA,
2009). The program has four overarching goals related
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 154
-------
to improving the reliability of New York's energy sys-
tem, reducing the energy costs for New Yorkers, miti-
gating environmental and health effects associated with
energy use, and creating economic benefits for the state.
The Energy Smart Program (E$P) is funded by a
System Benefits Charge (SBC) on the state's investor-
owned utilities and, since 1998, New York has spent
more than $1 billion to support it. The program's suc-
cess and broad impact are products of a commitment
to comprehensive evaluation, objective analysis, and
collaboration in order "to ensure that the successes and
failures of diverse programs are accurately and appro-
priately measured and reported" (NYSERDA, 2006b).
As part of that comprehensive evaluation process,
NYSERDA produces an annual report detailing the
multiple benefits of E$P on both a retrospective and
prospective basis. NYSERDA recognizes that program
expenditures "have substantial macroeconomic im-
pacts that go beyond these direct benefits" because the
"...purchase of goods and services through the Pro-
gram set off a ripple effect of spending and re-spending
that influences many sectors of the New York economy,
and the level and distribution of employment and
income in the State" (NYSERDA, 2009). NYSERDA
therefore conducts a periodic macroeconomic impact
analysis to quantify the full range of macroeconomic
impacts, expressed in terms of net annual employment,
labor income, total industrial output, and value added.
Method(s) Used
For the 2009 analysis, NYSERDA used the REMI Policy
Insight model, a macroeconomic model that combines
elements of input-output, econometric, and computable
general equilibrium models, to conduct the analysis.
New York estimated the positive and negative direct
effects of the program associated with the program's
expenditures and associated energy savings. These ef-
fects include:
an increase in demand for clean energy-related
goods and services,
an increase in disposable income for residential
customers due to the energy savings,
a reduction in productivity costs for business cus-
tomers whose energy costs have been reduced as a
result of the programs,
a decrease in disposable income for residents from
paying the SBC,
an increase in production costs for businesses from
paying the SBC charge, and
an increase in costs to residents and businesses
from purchasing the clean-energy-related goods
and services.
The data necessary to determine these effects have been
collected since E$P was implemented in 1999.
The analysis estimates historical macroeconomic
impacts of the program from 1999 through 2008, and
projects future impacts through 2022, assuming the
program funding ends in 2008.
Results
The results of the macroeconomic impact analysis indi-
cated that E$P has provided and will continue to pro-
vide net benefits in the form of increased employment,
personal income, total output, and gross state product.
The model indicated that E$P initiatives implemented
from 1999 through 2008 have already created 4,900 net
jobs across the following sectors:
2,134 jobs in the Personal and Business Services
sector,
841 in the Wholesale and Retail Trade sector,
794 in the Construction sector,
586 in the Transportation-related sector,
359 in State and Local Government, and
186 in Manufacturing.
During the same time period, the model showed that
the program increased personal income by $293 mil-
lion, gross state product by $644 million, and total
output by $1 billion.
The model was used to estimate the cumulative results
projected out to 2020, assuming that funding stops in
2008. During this 24-year period, E$P is expected to:
Create 86,400 net job years,
Increase personal income by $5.75 billion,
Increase gross state product by $13.37 billion, and
Increase total output by $20.59 billion (NYSERDA,
2009).
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 155
-------
NYSERDA evaluates E$P's macroeconomic impacts,
as well as the energy system and environmental and
health benefits, as part of its ongoing and comprehen-
sive evaluation strategy. The E$P program analyses
provide support for further development and imple-
mentation of clean energy initiatives. NYSERDA also
collaborates with independent parties, partners with
other government entities, and integrates its analyses
into the public policy forum via a 24-member advisory
group. NYSERDAs program underscores the impor-
tance of fully accounting for the multiple benefits of
clean energy initiatives in establishing the basis for
investment in energy efficiency and renewable energy.
For More Information
New York Energy $martSM Program Evalua-
tion and Status Report. NYSERDA. Report to
the System Benefits Charge Advisory Group.
May, 2006. http://www.nyserda.org/Energy_
Information/06sbcreport. asp.
5.3.2 ILLINOIS: ANALYZING THE
MACROECONOMIC BENEFITS OF CLEAN
ENERGY DEVELOPMENT
Benefits Assessed
Jobs
Household income
Business income
Clean Energy Program Description
In July 2005, the Illinois Commerce Commission voted
to adopt a Sustainable Energy Plan, the culmination
of years of work by the governor's Special Task Force
on the Condition and Future of the Illinois Energy
Infrastructure. The initial Sustainable Energy Plan (the
"Plan") proposal included provisions for both renew-
able energy portfolio standards and energy efficiency
portfolio standards, specifically:
A Renewable Portfolio Standard (RPS) that re-
quired an increasing percentage of electricity sold
to Illinois customers generated by renewable re-
sources: 2 percent by 2006, and increasing annually
by 1 percent until 2012.
The RPS further stipulated, as determined by the
study, that 75 percent of the renewable generation
should come from wind resources.
An Energy Efficiency Portfolio Standard (EEPS)
that required electricity load growth to be reduced
by the following amounts each year: 10 percent of
projected load growth in 2006-2008, 15 percent
in 2009-2011, 20 percent in 2012-2014, and 25
percent in 2015-2017.
The Illinois Commerce Commission's decision to adopt
the Plan followed more than five months of public
comment and deliberation among many stakeholders,
including utility companies and public interest groups.
Ultimately, the decision was largely guided by the pro-
posed Plan's substantial benefits, which were quantified
in a study released by the Energy Resources Center at
the University of Illinois in June of 2005 (Bournakis
and Hewings et al., 2005).
Method(s) Used
The direct and indirect macroeconomic impacts of
the Plan's provisions were analyzed using the Illinois
Regional Economic Input-Output Model (ILREIM).9
ILREIM includes two components, an input-output
model and an econometric model. The model links the
regional input-output component with macroeconomic
and demographic variables in a dynamic framework
that is able to examine the feedback effects of economic
events with different sectors.
More specifically, this model is a system of linear
equations formulated to predict the behavior of 151
endogenous variables, and consists of 123 behavioral
equations, 28 accounting identities, and 68 exogenous
variables. The model identifies 53 industries and three
government sectors.
For each industry in the structure, the model projects
output, employment, and earnings. The model also esti-
mates GSP, personal consumption expenditures, invest-
ment, state and local government expenditures, exports,
labor force, unemployment rate, personal income, net
migration, population, and the consumer price index.
To run ILREIM, the researchers provided data de-
scribing the dollar value of energy savings, the actual
electricity savings, and the various investments needed
to support the RPS and EEPS described in the Plan. The
scenarios that were run included large investments in ef-
9 Looking at the models described in Table 5.2.4, ILREIM is more like REMI
than IMPLAN or RIMS II.
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 156
-------
ficiency equipment and large investments in renewable
generation facilities relative to the baseline scenario.
Results
The University of Illinois analysis found, among other
benefits, that by 2012 the Plan would:
create 7,800 jobs and
generate nearly $9 billion in additional household
and business income.
In addition, the study also revealed other results:
The state would experience an economic adjust-
ment composed of the interplay between the
reduced local production of fossil-fuel energy and
the increased production of efficiency equipment
(it is likely that some portions of the efficiency
equipment will be manufactured in Illinois).
Part of the saved energy will come from reduced
energy imports.
Non-local impacts will affect the economies of
other states.
As indicated in the Illinois Commerce Commission's
Resolution10 to adopt the Plan, the realization (within
the Commission and among interested stakeholders)
that the Plan would "lead to rural economic develop-
ment" and create other environmental benefits was a
key factor in the Plan's final implementation. Further-
more, the transparent, detailed, and comprehensive
nature of the benefits study assured that, even after an
extensive review and comment period, the Plan ulti-
mately adopted by the Commission was nearly identi-
cal to the governor's original proposal.
For More Information
The Economic and Environmental Impacts of Clean
Energy Development in Illinois. Bournakis, A., G.
Hewings, J. Cuttica, and S. Mueller. Submitted to
the Illinois Department of Commerce and Eco-
nomic Opportunity. June, 2005. http://www.erc.uic.
edu/PDF/Clean_Energy_Development.pdf.
10 ICC Resolution 05-0437, available at: http://www.dsireusa.org/docutnents/
Incentives/IL04R.pdf
Sampling of State Clean Energy Analyses by Type of Analytic Method
State-level Clean Energy Analyses that Used I-O Analyses
Grover, S. 2007. Economic Impacts of Oregon Energy Tax Credit Programs Oregon
(BETC/RETC). Prepared by ECONorthwest for the Oregon Department of Energy.
May.
Nayak, N. 2005. Redirecting America's Energy: The Economic and Consumer
Benefits of Clean Energy Policies. Prepared by the U.S. PIRG Education Fund.
February.
Pletka, R. 2004. Economic Impact of Renewable Energy in Pennsylvania.
Prepared by Black & Veatch for The Heinz Endowments and Community
Foundation for the Alleghenies. March.
RAP. 2005. Electric Energy Efficiency and Renewable Energy in New England: An
Assessment of Existing Policies and Prospects for the Future. Prepared by The
Regulatory Assistance Project and Synapse Energy Economics, Inc. May.
Stoddard, L, J. Abiecunas, and R. O'Connell. 2006. Economic, Energy, and
Environmental Benefits of Concentrating Solar Power in California. Prepared by
Black & Veatch for U.S. DOE National Renewable Energy Laboratory. April.
U.S. DOC. 2003. Developing a Renewable Energy Based Economy for South
Texas - A Blueprint for Development. U.S. Department of Commerce, Economic
Development Administration, and the University of Texas at San Antonio.
U.S.
Pennsylvania
New England
California
Texas
h ttp://www.oregon.gov/
ENERGY/CONS/docs/EcoNW_
Study.pdf
h ttp://newenergyfuture.
com/newenergy.
asp?id2=15905&id3=energy
http://www.bv.com/Downloads/
Resources/ 'Reports/ 'PA_ RPS_
Final_Report.pdf
http://www.raponline.org/Pubs/
RSWS-EEandREinNE.pdf
h ftp://www.nrel.gov/docs/
fy06osti/39291.pdf
http://www.solarsanantonio.
org/pdf/EDAReport.pdf
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 157
-------
Sampling of State Clean Energy Analyses by Type of Analytic Method
SWEEP. 2002. The New Mother Lode: The Potential for More Efficient Electricity Southwest
Use in the Southwest. Southwest Energy Efficiency Project, Report for the
Hewlett Foundation Energy Series. November.
h ttp://www.swenergy.org/nml
State- Level Clean Energy Analysis that Used Econometric Models
Bernstein, M., C. Pernin, S. Loeb, and M. Hanson. 2000. The Public Benefit of California
California's Investments in Energy Efficiency. Prepared by RAND Science and
Technology for California Energy Commission. March.
Bernstein, M., R. Lempert, D. Loughram, and D. Ortiz. 2002a. The Public Benefit Massachusetts
of Energy Efficiency to the State of Massachusetts. Prepared by RAND Science
and Technology.
Bernstein, M., R. Lempert, D. Loughram, and D. Ortiz. 2002b. The Public Benefit Minnesota
of Energy Efficiency to the State of Minnesota. Prepared by RAND Science and
Technology.
Bernstein, M., R. Lempert, D. Loughram, and D. Ortiz. 2002c. The Public Benefit Washington
of Energy Efficiency to the State of Washington. Prepared by RAND Science and
Technology for the Energy Foundation. February.
h ftp://rand.org/pubs/
monograph_reports/2005/
MR1212.0.pdf
http://www.rand.org/pubs/
monograph_reports/2005/
MR1588.pdf
http://www.rand.org/pubs/
monograph_reports/2005/
MR1587.pdf
http://www.rand.org/pubs/
monograph_reports/2005/
MR1589.pdf
State- Level Clean Energy Analysis that Used Computable General Equilibrium Models
Roland-Hoist, D. University of California Berkeley. Economic Assessment of some California
California Greenhouse Gas Control Policies: Applications of the BEAR Model. No
date given.
h ttp://calclima te. berkeley.
edu/research/ghg/assets/2_
Economic_Assessment.pdf
State- Level Clean Energy Analysis that Used Hybrid Models
Bournakis, A., G. Hewings, J. Cuttica, and S. Mueller. 2005. The Economic and Illinois
Environmental Impacts of Clean Energy Development in Illinois. Submitted to the
Illinois Department of Commerce and Economic Opportunity. June.
CT GSC. 2004. 2005 Climate Change Action Plan, Appendix 9: Economic Impact Connecticut
of Oil and Natural Gas Conservation Policies. Connecticut Governor's Steering
Committee, prepared by Regional Economic Models, Inc. November.
DOER. 2004. 2002 Energy Efficiency Activities. Massachusetts Division of Energy Massachusetts
Resources.
Hewings, G., and M. Yanai, 2002. Job Jolt: The Economic Impacts of Repowering Midwest
the Midwest. Prepared by the Regional Economics Applications Laboratory.
Jensen, V., and E. Lounsbury. 2005. Assessment of Energy Efficiency Potential Georgia
in Georgia. Prepared for the Georgia Environmental Facilities Authority by ICF
Consulting. May.
NYSERDA. 2009. New York Energy Smart Program Evaluation and Status Report; New York
Year Ending December 31, 2008, Report to the Systems Benefit Charge Advisory
Group, Final Report, March.
h ttp://www.erc. uic. edu/PDF/
Clean_Energy_Development.pdf
http://www.ctd/matechange.
com/documen ts/
Appendix9_REMI_HeatingOil
andNaturalGasConservation
Funds_CCCAP_2005_OOO.pdf
h ttp://www.mass.gov/Eoca/
docs/doer/pub_info/ee02-long.
pdf
http://www.issuelab.org/
research/job_jolt_ the_
economic_impacts_of_
repowering_ the_midwest
h ttp://www.gefa. org/
Modules/ShowDocument.
aspx?documentid=46
h ttp://www.nyserda. org/
publications/SBC%20March%20
2009%20Annual%20Reportpdf
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 158
-------
Sampling of State Clean Energy Analyses by Type of Analytic Method
Reference
Sumi, D., G. Weisbrod, B. Ward, and M. Goldberg. 2003. An Approach to
Quantifying Economic and Environmental Benefits for Wisconsin's Focus on
Energy. Presented at International Energy Program Evaluation Conference. August.
State/Region
Wisconsin
h ttp://edrgroup.com/pdf/sumi-
weisbrod-wis-energy-iepec.pdf
Weisbrod, G., K. Polenske, T. Lynch, and X. Lin. 1995. The Economic Impact
of Energy Efficiency Programs and Renewable Power for Iowa: Final Report.
Economic Development Research Group, Boston, MA. December.
Iowa
http://www.edrgroup.com/
library/energy-environment/
iowa-energy.h tml
Information Resources
URL Address
Energy 2020 model
ICF International Inc. IPM model
Minnesota IMPLAN Group, Inc. IMPLAN model
h ttp://www. energy2020. com/model_ overview.h tm
http://www.icfi.com/Markets/Energy/energy-modeling. asp#2
h ttp://www. implan.com/
Regional Economic Modeling, Inc. REMI model http://www.remi.com/
REPP Labor Calculator
Rocky Mountain Institute Community Energy Opportunity Finder tool
RTI International Applied Dynamic Analysis of the Global Economy
(ADAGE) model
U.S. Bureau of Economic Analysis Regional Economic Accounts
U.S. Census Bureau
U.S. Department of Commerce RIMS II model
U.S. DOEArgonne National Laboratory Long-Term Industrial Energy
Forecasting (LIEF) model
U.S. DOE Lawrence Berkeley National Laboratory DOE-2.2 model
U.S. DOE National Renewable Energy Laboratory Excelergy model
U.S. DOE National Renewable Energy Laboratory Jobs and Economic
Development Impact (JEDI) tool
U.S. Energy Information Administration
University of Georgia, Georgia Economic Modeling Systems (GEMS)
University of Illinois, Regional Economics Applications Laboratory (REAL)
h ttp://www. cres t.org/
http://www.energyfinder.org/
http://www.rti.org/page.cfm?objectid=DDC06637-7973-
4BOF-AC46B3C69E09ADA9
http://www.bea.gov/regional/index.htm
http://www.census.gov/
h ttps://www. bea.gov/regional/rims/
http://www.dis.anl.gov/projects/EnergyAnalysisTools.
htmimief
h ttp://www. doe2. com/
h ftp://www.nrel.gov/
http://www.nrel.gov/analysis/jedi/
h ttp://www. eia. doe.gov/
h ttp://www. cviog. uga. edu/
h ttp://www. real. uiuc. edu/
References
Bernstein, M., C. Pernin, S. Loeb, and M. Hanson. 2000. The Public Benefit
of California's Investments in Energy Efficiency. Prepared by RAND Science
and Technology for California Energy Commission. March.
URL Address
http://rand.org/pubs/monog raph_reports/2005/
MR1212.0.pdf
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 159
-------
References
Bernstein, M., R. Lempert, D. Loughram, and D. Ortiz. 2002a. The Public
Benefit of Energy Efficiency to the State of Massachusetts. Prepared by
RAND Science and Technology.
Bernstein, M., R. Lempert, D. Loughram, and D. Ortiz. 2002b. The Public
Benefit of Energy Efficiency to the State of Minnesota. Prepared by RAND
Science and Technology.
Bernstein, M., R. Lempert, D. Loughram, and D. Ortiz. 2002c. The Public
Benefit of Energy Efficiency to the State of Washington. Prepared by
RAND Science and Technology for the Energy Foundation. February.
h ttp://www. ran d. org/p ubs/m on ograph_ reports/20 05 /
MR1588.pdf
h ttp://www. ran d. org/p ubs/m on ograph_ reports/20 05 /
MR1587.pdf
h ttp://www. ran d. org/p ubs/m on ograph_ reports/20 057
MR1589.pdf
Bohringer, C. 1998. The Synthesis of Bottom-Up and Top-Down in Energy http://www.elsevier.com/wps/find/journaldescription. cws_
Policy Modeling. Energy Economics, v. 20, pg. 223-248. home/30413/description#description
Bournakis, A., G. Hewings, J. Cuttica, and S. Mueller. 2005. The Economic
and Environmental Impacts of Clean Energy Development in Illinois.
Submitted to the Illinois Department of Commerce and Economic
Opportunity. June.
Cinnamon, B., T. Beach, M. Huskins, and M. McClintock. 2005. The
Economics of Solar Power for California. August.
CT GSC. 2004. 2005 Climate Change Action Plan, Appendix 9: Economic
Impact of Oil and Natural Gas Conservation Policies. Connecticut
Governor's Steering Committee, prepared by Regional Economic Models,
Inc. November.
Department of Administration (DOA), State of Wisconsin. 2001. Focus
on Energy Pilot Study: Demonstration of Economic Impact Analysis for
Commercial & Industrial Programs. Prepared by Economic Development
Research Group, Inc. September.
DOER. 2004. 2002 Energy Efficiency Activities. Massachusetts Division of
Energy Resources.
Goldberg, M., K. Sinclair, and M. Milligan. 2004. Job and Economic
Development Impact (JEDI) Model: A User-Friendly Tool to Calculate
Economic Impacts from Wind Projects. Presented at the 2004 Global
WINDPOWER Conference, Chicago, Illinois. March.
Grover, S. 2007. Economic Impacts of Oregon Energy Tax Credit Programs
in 2006 (BETC/RETC). Prepared by ECONorthwest for the Oregon
Department of Energy. May.
Heavner, B. and B. Del Chiaro. 2003. Renewable Energy and Jobs:
Employment Impacts of Developing Markets for Renewables in California.
Environment California Research and Policy Center. July.
Hewings, G., and M. Yanai, 2002. Job Jolt: The Economic Impacts
of Repowering the Midwest. Prepared by the Regional Economics
Applications Laboratory.
IREC. 2005. Labor Forecasts and Job Trends for the Solar and Renewable
Energy Industries. Interstate Renewable Energy Council Notebook,
Version #5. January.
Jensen, V., and E. Lounsbury. 2005. Assessment of Energy Efficiency
Potential in Georgia. Prepared for the Georgia Environmental Facilities
Authority by ICF Consulting. May.
h ftp://www.erc. uic. edu/PDF/Clean_ Energy _ Developmen t.
pdf
http://akeena.net/Library/pdf s/Economics_of_Solar_
Power_for_California_Aug2005.pdf
http://www.ctclimatechange.com/documents/Appendix9_
REMI_HeatingOilandNaturalGasConservationFunds_
CCCAP_2005_OOO.pdf
http://www.edrgroup.com/pdf/Wis-Pilot-Report.pdf
http://www.mass.gov/Eoca/docs/doer/pub_info/ee02-long.
pdf
http://www.nrel.gov/docs/fy04osti/35953.pdf
http://www.oregon.gov/ENERGY/CONS/docs/EcoNW_
Study.pdf
http://www.environmentcalifornia.org/uploads/OW/aa/
OWaa2RaedlfHwQOWbxKd5w/Renewable_Energy_and_
Jobs.pdf
http://www.issuelab.org/research/job_jolt_the_economic_
impacts_of_repowering_the_midwest
h ftp://www.irecusa.org/index.php ?id= 6
www.gefa.org/Modules/ShowDocument.
aspx?documentid=46
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 160
-------
Loschel, A. 2002. Technological Change in Economic Models of
Environmental Policy: A Survey. Ecological Economics, v. 43
Mongha, N., E. Stafford, and C. Hartman. 2006. An Analysis of the
Economic Impact on Utah County, Utah from the Development of Wind
Power Plants. Renewable Energy for Rural Economic Development, Utah
State University. DOE/GO-102006-2316. May.
MTC. 2005. Energy Efficiency, Renewable Energy, and Jobs in
Massachusetts. Massachusetts Technology Collaborative, Renewable
Energy Trust. November.
Nayak, N. 2005. Redirecting America's Energy: The Economic and
Consumer Benefits of Clean Energy Policies. Prepared by the U.S. PIRG
Education Fund. February.
Nevada AFL-CIO. Comments Submitted to the Nevada Public Service
Commission: Procedural Order No. 3 and Request for Comments No. 2.
New York Office of the Governor. 2003. Press Release: Governor
Announces $14.5 Million for Clean Energy Projects. July 18, 2003.
NJ BPU. 2005. New Jersey's Clean Energy Program: 2005 Annual Report.
New Jersey Board of Public Utilities, Office of Clean Energy.
NYSERDA. 2006a. Press Release: State Energy Authority Announces
Partnership that will Lead to the Growth of Clean-Energy and
Environmental Technologies in New York State. July 5.
NYSERDA. 2006b. New York Energy Smart Program Evaluation and Status
Report: Report to the Systems Benefits Charge Advisory Group. May.
NYSERDA. 2008. New York Energy Smart Program Evaluation and Status
Report; Year Ending December 31, 2007, Report to the Systems Benefit
Charge Advisory Group, Final Report, March.
NYSERDA. 2009. New York Energy Smart Program Evaluation and Status
Report; Year Ending December 31, 2008, Report to the Systems Benefit
Charge Advisory Group, Final Report, March.
Pletka, R. 2004. Economic Impact of Renewable Energy in Pennsylvania.
Prepared by Black & Veatch for The Heinz Endowments and Community
Foundation for the Alleghenies. March.
RAP. 2005. Electric Energy Efficiency and Renewable Energy in New
England: An Assessment of Existing Policies and Prospects for the Future.
Prepared by The Regulatory Assistance Project and Synapse Energy
Economics, Inc. May.
REMI. 2004. Economic Impact of Oil and Natural Gas Conservation
Policies. Prepared for U.S. Environmental Protection Agency and the State
of Connecticut. November.
REPP. 2005. Component Manufacturing: Ohio's Future in the Renewable
Energy Industry. Renewable Energy Policy Project. October.
Roland-Hoist, D. 2006. University of California Berkeley. Economic
Assessment of some California Greenhouse Gas Control Policies:
Applications of the BEAR Model. August.
http://www.elsevier.com/locate/ecolecon
h ttp://www. eere.en ergy. gov/win dan dhydro/
windpoweringamerica/pdfs/wpa/econ_dev_jedi.pdf
http://www.masstech.org/IS/reports/clusterreportll405.pdf
http://newenergyfuture.com/newenergy.
asp?id2=15905&id3=energy
http://www.repp.org/articles/static/l/binaries/Nevada_RPS.
pdf
h ttp://www.nyserda. org/Press_ Releases/press_
a rch ives/2003/govern or/govjulyl 8_ 03. asp
http://www.njcleanenergy.com/files/file/2005%20
Annual%20Reportpdf
http://www.nyserda.org/Press_Releases/2006/
PressRelease20060507.asp
http://www.nyserda.org/Energy_lnformation/06sbcreport.
asp
http://www.nyserda.org/pdfs/Combined%20Report.pdf
http://www.nyserda.org/publications/SBC%20March%20
2009%20Annual%20Reportpdf
http://www.bv.com/Downloads/Resources/Reports/PA_
RPS_Final_Report.pdf
http://www.raponline.org/Pubs/RSWS-EEandREinNE.pdf
http://ctclimatechange.com/documents/Appendix9_
REMI_HeatingOilandNaturalGasConservationFunds_
CCCAP_2005_OOO.pdf
http://www.repp.org/articles/static/l/binaries/Ohio_
Manufacturing_Report_2.pdf
http://calclimate.berkeley.edu/research/ghg/assets/2_
Economic_Assessment.pdf
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 161
-------
URL Address
Roland-Hoist, D. Economic Growth and Greenhouse Gas Mitigation in
California. August 2006. UC Berkeley
http://utahcleanenergy.org/files/ul/d_GHG_Mitigation_in_
CA_Exec_Summ_UC_Berkeley.pdf
Spitzer, E. 2007. "15 by 15" A Clean Energy Strategy for New York. Speech
by former Governor Spitzer, April 19.
h ttp ://www.state.ny.us/governor/keydocs/0419071_ speech.
html
Stoddard, L, J. Abiecunas, and R. O'Connell. 2006. Economic, Energy,
and Environmental Benefits of Concentrating Solar Power in California.
Prepared by Black & Veatch for U.S. DOE National Renewable Energy
Laboratory. April.
http://www.nrel.gov/docs/fy06osti/39291.pdf
Sue Wing, I. 2004. Computable General Equilibrium Models and Their Use
in Economy-Wide Policy Analysis: Everything You Ever Wanted to Know
(But Were Afraid to Ask). MIT Joint Program on the Science and Policy of
Global Change. Technical Note No. 6, September.
http://mit.edu/globalchange/www/MITJPSPGC_TechNote6.
pdf
Sumi, D., G. Weisbrod, B. Ward, and M. Goldberg. 2003. An Approach
to Quantifying Economic and Environmental Benefits for Wisconsin's
Focus on Energy. Presented at International Energy Program Evaluation
Conference. August.
http://edrgroup.com/pdf/sumi-weisbrod-wis-energy-iepec.
pdf
SWEEP. 2002. The New Mother Lode: The Potential for More Efficient
Electricity Use in the Southwest. Southwest Energy Efficiency Project,
Report for the Hewlett Foundation Energy Series. November.
h ftp://www.swenergy.org/nml
U.S. DOC. 2003. Developing a Renewable Energy Based Economy
for South Texas - A Blueprint for Development. U.S. Department of
Commerce, Economic Development Administration, and the University of
Texas at San Antonio.
http://www.solarsanantonio.org/pdf/EDAReport.pdf
U.S. EPA. 2006. National Action Plan for Energy Efficiency. Office of Air
and Radiation, Climate Protection Partnerships Division. July.
h ttp://www.epa.gov/cleanenergy/documen ts/napee/
napee_report.pdf
Weisbrod, G., K. Polenske, T. Lynch, and X. Lin. 1995. The Economic
Impact of Energy Efficiency Programs and Renewable Power for Iowa:
Final Report. Economic Development Research Group, Boston, MA.
December.
http://www.edrgroup.com/library/energy-environment/
iowa-energy.h tml
Wisconsin Department of Administration. 2007. Division of Energy. Focus
on Energy Public Benefits Evaluation. Economic Development Benefits:
FY07 Economic Impacts Report. Final: February 23, 2007.
http://www.focusonenergy.com/data/common/dmsFiles/E_
EC_ RPTI_Econ_ Dev_ Benefits_ FY07.pdf
CHAPTER 5 | Assessing the Multiple Benefits of Clean Energy 162
-------
APPENDIX A
Catalogue of Clean Energy
Case Studies Highlighted in
the Multiple Benefits Guide
U
o
Q
CHAPTER ONE
Introduction
CHAPTER TWO
Potential Energy Impacts of Clean Energy
CHAPTER THREE
Electric System Benefits of Clean Energy
CHAPTER FOUR
Air Quality Benefits of Clean Energy
CHAPTER FIVE
Economic Benefits of Clean Energy
(|> APPENDIX A
Catalogue of Clean Energy Case Studies
APPENDIX B
Tools and Models Referenced in Each Chapter
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 163
-------
CATALOGUE OF CLEAN ENERGY CASE STUDIES HIGHLIGHTED IN THE MULTIPLE BENEFITS GUIDE
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Emissions, Air
Energy Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Connecticut:
Incorporating
Multiple Benefits
in Evaluation
Criteria for
New Capacity
Additions
Ohio: Clean
Energy Initiatives
Can Benefit
Economic
Development
In June 2005, Connecticut policymakers enacted
Public Act 05-01, An Act Concerning Energy
Independence (EIA), which authorized the
Connecticut Department of Public Utility Control to
launch a competitive procurement process geared
toward motivating new supply-side and demand-
side resources.
As part of the bid evaluation process, each capacity
project is scored based on a multiple benefits
weighting system: A total of 85% of the evaluation
score is based on a benefit-cost analysis of the
project. A total of 15% of the evaluation score is
determined through the assessment of five other
criteria with their associated weights (see benefits,
right).
A 2007 study by the American Solar Energy
Society assessed the renewable energy and
energy efficiency market and developed forecasts
of the market's future economic growth. The
study established a baseline of 2006 and forecast
the growth of the renewable energy and energy
efficiency industry from this baseline to 2030 under
three different scenarios. Using this approach,
the authors developed a case study for Ohio, an
area hard hit by the loss of manufacturing jobs.
The analysis concluded that the energy efficiency
and renewable energy industries offer significant
development opportunities in the state.
Connecticut Climate
Change 2005.
Connecticut Climate
Action Plan.
Bezdek. Roger.
2007. Renewable
Energy and Energy
Efficiency: Economic
Drivers for the 21st
Century. Prepared for
the American Solar
Energy Society.
Use of existing sites Reduced emissions Front-loading of costs - 2005-
and infrastructure of SO2, NOx, and 2.5% 2020
- 2.5% C02 - 5%
Benefits of fuel
diversity - 2.5%
Other benefits
(e g , transmission
reliability,
employment effects,
benefits of high level
efficiency such as
CHP) - 2.5%
In 2030: 2006
^,0,.,,. . 2030
V.LO billion in r6V6nu6s
and 175,000 jobs
cinnLicilly in tns
Fen6Wฃibl6 en6rcfy
industry
$200 billion in
revenues and more
than 2 million jobs in
the energy efficiency
industry
P
P
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 164
-------
Chapter 1: Introduction
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Multiple States:
A recent study by the Lawrence Berkeley National
Reducing Laboratory (LBNL) examined several studies of the
Natural Gas
Prices through
Increased
natural gas consumer benefits from clean energy
programs. Most of the studies evaluated a national
or state RPS, or a combined RPS and EE program and
Deployment consistently showed that "RE and EE deployment
of Renewable will reduce natural gas demand, thereby putting
Energy and downward pressure on gas prices" (Wiser et al,
Energy Efficiency 2005).
Multiple States:
In 2004 the University of California-Berkeley
How Many Jobs reviewed 13 independent reports and developed a
Can the Clean model to examine the job creation potential of the
Energy Industry renewable energy industry.
Generate?
The study analyzed the employment implications of
three national 20% RPS scenarios and two scenarios
where the generation required by the RPS is
produced instead by fossil-fuel generation.
Wiser. R.. M. Bolinaer.
and M. Clair. 2005.
Easing the Natural
Gas Crisis: Reducing
Natural Gas Prices
through Increased
Deployment of
Renewable Energy
and Energy Efficiency.
Ernest Orlando
Lawrence Berkeley
National Laboratory
(LBNL). January.
Kammen. D.. K.
Kapadia. M. Fripp.
2004. Putting
Renewables to Work:
How Many Jobs Can
the Clean Energy
Industry Generate?
April.
Each 1% reduction in
2003-
national gas demand is 2020
likely to lead to a long-
term average reduction
in wellhead gas prices
of 0.8% to 2%.
The present value of
natural gas bill savings
from 2003-2020 are
within the range of $10
^4D hillinn
- VT-U UUUUll.
Consumers' gas bill
savings are estimated
between $7.50 and
$20 for each MWh of
electricity produced by
RE or saved with EE.
The RE industry
1998-
generates more jobs 2004
than the fossil-fuel
industries per unit of
energy delivered and
per dollar invested,
driven primarily by
the general shift from
mining and related
services to increased
manufacturing,
construction, and
installation activity.
P
R
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 165
-------
Chapter 1: Introduction
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
New England
and NE Canada:
Multiple Benefits
The Conference of New England Governors and
Eastern Canadian Premiers (NEG-ECP) developed a
comprehensive Climate Change Action Plan in 2001
Analysis is Being with the long-term goal of reducing GHG emissions
Used in Regional in the region by 75-85% and enacted Policy
Planning Resolution 30-2 to promote energy efficiency and
renewable energy in the region.
A study, Electric Energy Efficiency and Renewable
Energy In New England: An Assessment of Existing
Policies and Prospects for the Future, estimates that
by 2010, the combined effect of expected energy
efficiency and renewable energy deployment will
provide a wide range of benefits that go beyond
direct energy savings.
New England
Governors and
Eastern Canadian
Premiers (NEG-ECP).
2006. Resolution
30-2: Resolution
Concerning Energy.
May.
Energy security
Estimated
Estimated economic
2000-
benefits between environmental 1 benefits between 2000 2010
2000 and 2010 benefits between and 2010 included:
included:
2000 and 2010 . . . ^ ,
A npr nn^im/p Sn 1
A i. i ! i included: , .. , ^, . ,
A stabilizing and billion for the New
reducing influence savings of 31.6 England economy
on the wholesale million tons of .. .. 00^' i_
_ _ . . More than 28,000 job-
price of, and CO2 emissions
demand for, natural Oo^t f
22 000 tons of
gas . ' . . $1 billion in wages
NUX emissions
Reduced wholesale
. . . .. . . 34,000 tons of
electricity prices in _ . .
.. . SO2 emissions
the regional market
Reduced demand
for new facilities in
the electric market
Increased resiliency
of the grid
RandP
Chapter 2: Assessing the Potential Energy Impacts of Clean Energy Initiatives Programs
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
New Jersey:
Basic Demand
Forecast
The New Jersey energy plan-basic demand forecast
for 2005-2020. This study illustrates an example of a
linear extrapolation analysis.
The BAU electricity forecast was developed using a
relatively simple approach in which past load growth
rates were reviewed and assumptions were made
regarding the ways in which industry trends and
existing policies affect future growth patterns.
Summit Blue
Consulting 2008
Assessment of
the New Jersey
Renewable Energy
Market, Volume I and
II. Prepared for the
New Jersey Board of
Public Utilities. March.
The electricity
sectors from 2005-
2020 is projected to
be 1.52%
2005-
2020
P
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 166
-------
Chapter 2: Assessing the Potential Energy Impacts of Clean Energy Initiatives Programs
Case Studies
Key Benefits Findings, Results and Activities
State/Region
and Name of
Program
Summary of Policy/Program and Analysis
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
New York:
Energy SmartSM
Public Benefits
Program
New York's public benefit program was established
NYSERDA. 2005.
by order of the New York State Public Service New York Energy
Commission (PSC) in January 1998 and funded SMARTSM Program.
by the System Benefits Charge (SBC). New York Evaluation and Status
State Energy Research and Development Authority
(NYSERDA) administers the New York Energy
SmartSM Program which promotes competitive
markets for energy efficiency services, provides
Texas:
direct benefits to electricity ratepayers and/or to
the people of New York and stimulates demand
for energy-efficient products and services, and
renewable resource technologies.
In this study, NYSERDA uses a production costing
model, MAPS, to forecast the avoided energy and
capacity benefits of the programs for several years.
The legislation (Senate Bill 5, 2001) that initiated
Building Code the Texas Emissions Reduction Plan (TERP) requires
the Energy Systems Laboratory (ESL) at the Texas
Engineering Experiment Station of the Texas A&M
University System to submit an annual report to
the Texas Commission on Environmental Quality
estimating the historical and potential future energy
Report for the Year
Ending December
2004. New York Public
Service Commission
and New York State
Energy Research
and Development
Authority. May.
NYSERDA. 2008.
New York Energy
SMARTSM Program.
Evaluation and Status
Report for the Year
Ending December
2007. New York Public
Service Commission
and New York State
Energy Research
and Development
Authority. March.
Texas A&MEnergy
Systems Laboratory
(ESL). 2007. Energy
Efficiency/Renewable
Energy Impact in
the Texas Emissions
Reduction Plan
savinas from enerav buildina code adoption and. (TERP). Volume II-
when applicable, from more stringent local codes or
above-code performance ratings.
Using data from the TCEQ and EPA, including
eGRID, ESL estimated the energy savings and NOX
reductions from energy code compliance in new
residential construction. ESL has conducted this
annual analysis since 2002.
Technical Report.
Electricity savings of:
Reduced nearly
Between 1998-2004:
l,400GWh between f 60ฐ f??L4'70ฐ Saved $195 million in
.__ . tons of NO and
1998-2004 0_ * . energy costs
SOx respectively
3,000 GWh savings 1 Reduced annual energy
1 1 lArr^acAn anm lal
by 2007
Annual Energy
Savings:
1,440,885 MWh
of electricity each
" " bills by $570 million
CO, emissions by _ . , _,,..,
2.... Created and retained
2 million tons
4,700 jobs
NOX emissions
reduced by:
1,014 tons-NO/
year in 2007
year 2,047 tons/year by
Approximately 2.9 2013
million MWh by
2013
By 2027 the program is
expected to:
Create more than 7200
jobs
Increase labor income
more than $300 million
each year
Increase total annual
output in the state by
$503 million
1998-
2027
1998-
2027
RandP
P
' P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 167
-------
Chapter 2: Assessing the Potential Energy Impacts of Clean Energy Initiatives Programs
Case Studies
Key Benefits Findings, Results and Activities
State/Region
and Name of
Program
Summary of Policy/Program and Analysis
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Vermont:
Energy and
Energy Savings
Forecasting
Wisconsin:
The Vermont Department of Public Service (DPS)
conducts forecasting as a part of its long-term state
energy policy and planning process. The state
uses the Comprehensive Energy Plan (CEP) to help
manage the transition from fossil fuels to cleaner
energy in order to benefit Vermont's economic and
environmental future.
An analysis performed in 2008 inlcuded an
examination of historical energy consumption
across all sectors 1960-2005. The forecasting
process required the following steps: 1) Determine
fuel price projections and avoided costs; 2) Estimate
the achievable, cost-effective potential for electric
energy and peak demand savings; 3) Develop a 20-
year forecast of electric energy use; and 4) Develop
a peak demand forecast. It also employed historical
data to compare energy demand in Vermont with
New England from 1990-2004.
In 2006, then-Wisconsin Governor Jim Doyle
Office of Energy launched the Declaration of Energy Independence,
Independence:
Vermont's Energy
Forecasting Efforts.
Vermont Department
of Public Service.
June 19. 2008
Wisconsin Office of
Energy Independence.
which included a goal of using renewable energy to 2007. Wisconsin
Demand & generate 25 percent of the state's electricity and 25
supply baselines percent of its transportation fuels by 2025. It uses
& energy a top-down approach to help a state understand
consumption by the large and small consumers within the state and
fuel type data helps target sectors for policy interventions. It also
employs a bottom-up approach to explore a sector-
or technology specific clean energy policy.
This analysis was performed in 2007 by breaking
down consumption data by the sectors that
consume the fuels, including the commercial,
residential, industrial, transportation, and utility
sectors. Consumption and/or generation-related
baseline data can be obtained from DOE's EIA, EPA's
eGRID, NERC, lOSs, public utility commissions, and
many more.
Energy Statistics.
Electricity demand
is expected to grow
an average of 0.93%
on an average
annual basis 2008-
2028.
When new DSM
measures are
implemented,
DPS anticipates a
decline of 0.19% on
an average annual
basis.
Overall petroleum
From 2008-2009
use decreased 2.3%
in 2009. Of the y . 2
6missions
total petroleum decreased 18 9
Due to forecasts of a
1960
large supply gap with 2005
high costs to replace
power contracts,
Vermont committed
itself to pursue very
aggressive energy
efficiency measures.
Total electricity sales
decreased 6.4% in 2009
but have grown 3.7%
over the past ten years.
used in Wisconsin, ,.,,,.,> , i .. -t i
_, . ... percent Utility In 2009, electricity sales
81.4 percent is in r\_ . . , , . .
NO emissions decreased in all sectors
the transportation , x , 00 0
decreased 28 2
sector, which saw a t
decrease of 4.2%.
perceru
1970
2006
P
R
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 168
-------
Chapter 3: Assessing the Electric System Benefits of Clean Energy Programs
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
California:
The California Public Utilities Commission (CPUC)
Utilities' Energy approved a new method for calculating avoided
Efficiency costs for use in evaluating utility energy efficiency
Programs
Energy Efficiency
Portfolio Plans and
Program Funding
programs in California and demonstrated how clean Levels for 2006-
energy can be used in the state energy planning and
policy decision-making process.
A new methodology was used that includes five
major categories of costs that are avoided when
demand is reduced through installation of energy
efficiency resources. It produces time- and location-
specific cost estimates, whereas the previous
avoided-cost methodology relied more upon
Massachusetts:
Energy Efficiency
average statewide values.
This study explores the potential price and emissions
benefits of increasing distributed generation,
and Distributed photovolatics (PV), combined heat and power (CHP)
Generation and energy efficiency in Massachusetts through
2020.
A reference case was developed to determine what
the wholesale electric prices and carbon dioxide
emissions would be without the additional clean
energy resources. PROSYM simulation model was
used to determine the potential price and emissions
impacts of the four scenarios which are then
compared against the reference case to determine
the impacts.
2008-Phase 1 Issues.
California Public
Utilities Commission
Interim Opinion
September 22. 2005.
Impacts of Distributed
Generation on
Wholesale Electric
Prices and Air
Emissions in
Massachusetts.
Synapse Energy
Economics. March 51.
2008.
The electricity
growth rate for all
sectors from 2005-
2020 is projected to
be 1.52%
Each scenario was
found to achieve
Avoided electricity
2006
generation costs: 2008
$133/MWh with
the new method
(compared with $807
MWh with the old
method)
Avoided T&D costs:
Avoided environmental
externality costs
Avoided ancillary
services costs
Reduced wholesale
market clearing prices
The 250MW of PV is
expected to displace
reductions of CO2 356 GW of purchases
emissions relative from the wholesale
to the reference market and reduce
case: EE and CHP prices by 0.4%
combined will __ . .
. EE is expected to
have a reduction . . ,.,.,,
, . .... reduce prices by 1.6%
of 2 4 million
short tons CO2/
year in 2020
EE and CHP would
produce 5.1% reduction
2007
2020
P
P
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 169
-------
Chapter 3: Assessing the Electric System Benefits of Clean Energy Programs
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Northeast:
In all four of the structured, RTO-run eastern spot
Price Effects electricity markets, historically high peak load values
of Demand occurred during a week-long heat wave in August
Response 2006. Market coordinators all acknowledged the
Vermont:
System Planning
Approach to
Estimate Avoided
Transmission
Costs
role that demand response (DR) played in keeping
peak load lower than what otherwise would have
occurred and the study estimate the wholesale price
effects from using DR during these peak times.
Vermont: System Planning Approach to Estimate
Avoided Transmission Costs The Vermont Electric
Company (VELCO) undertook a study in 2003 of
alternatives to a proposed major upgrade in the
northwest corner of Vermont. VELCO conducted
a thorough study of distributed generation, energy
efficiency, and new central generation as alternatives
to the upgrade. It demonstrates one way to use
the system planning approach to estimate avoided
transmission costs.
The study identified a range of central generation
and distributed generation options and estimated
their costs. In addition, a location -specific study of
the available energy efficiency potential and the
program costs for delivering that potential was
prepared. Various combinations of energy efficiency
and generation were assembled as alternatives to
the proposed transmission project and compared
based on total present value of cost of service.
"Early Aug. Demand
Response Produces
$650 Million Savings
in PJM: Reducing
Electricity Use
Stretches Power
Supplies. Lowers
Wholesale Electricity
Supplies." August 17.
2006.
LaCapra Associates.
2005. Alternatives To
Velco's Northwest
Vermont Reliability
Project. January 29
(LaCapra Associates.
2003: Orans. 1989:
Orans. 1992).
Wholesale prices would
have been $300/MWh
higher without demand
response during heat
wave
Demand response to
heat wave reported
savings of about $650
million for energy
purchasers
The study determined
the cost of transmission
upgrade and the cost
of a smaller upgrade;
the difference in those
two costs could be
used to assess the
cost-effectiveness of
the alternative resource
package
2006
2002-
2011
R
P
* P = Prospective: R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 170
-------
Chapter 4: Air Pollution, Greenhouse Gas, Air Quality & Health Benefits of Clean Energy Programs
Case Studies
Key Benefits Findings, Results and Activities
State/Region Time
and Name of Emissions, Air Period for Type of
Program Summary of Policy/Program and Analysis Link Energy Quality, and Health Economic Analysis Analysis*
Connecticut:
Connecticut worked with EPA and NESCAUM to
Economic quantify the economic, air quality, and health
Impact of Oil benefits of policy options while developing the
and natural Gas state's 2005 Climate Change Action Plan. The
Conservation state specifically analyzed the benefits of oil and
Connecticut GSC
on Climate Change.
2005. CCCAP. GSC
on Climate Change.
Connecticut Climate
and results of natural gas conservation programs that encourage Change Web site.
using COBRA
installation of EE equipment. Three scenarios
analyzed from 2005-2020: oil program, gas
program, combined programs. Program funded by
a 3% natural gas and oil-use surcharge. Emissions
were estimated through the development of their
Climate Change Action Plan. Macroeconomic effects
modeled with REMI. Public health effects from
State Action Plan.
CT GSC. 2004. 2005
Climate Change
Action Plan. Appendix
9: Economic Impact
of Oil and Natural
avoided emissions estimated with EPAs COBRA Gas Conservation
mocjel Policies. Connecticut
'
Governor's Steering
Committee, prepared
by Regional Economic
Models. Inc.
November.
By 2020:
Net benefits from 2005-
2005
2020 include ($1996): 2020
Ou programs
are expected 2,092 average annual
to avoid: 1.89 jobs
millions of metric ,*,
, , $3.1M output
tons of carbon
dioxide equivalent $2.03M GSP
(MMTCO,e) .... ... , ,. , .
2 $1.8M real disposable
Gas programs are income
expected to avoid:
2.07MMTC02e.
An additional $4 to $1
payback of reduced
health costs and public
health benefits was
identified as a result of
reductions in criteria air
pollutants.
P
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 171
-------
Chapter 4: Air Pollution, Greenhouse Gas, Air Quality & Health Benefits of Clean Energy Programs
Case Studies
Key Benefits Findings, Results and Activities
State/Region
and Name of
Program
.ary of Policy/Program and Analysis
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Minnesota:
For testimony to the Minnesota Public Utilities
How BenMAP Commission about building a new clean energy
Has Been electricity generating facility, Excelsior Energy
Used in Clean
Energy Analysis:
compared the air quality and health effects of two
proposed 600 MW integrated gasification and
Excelsior Energy.
2005. Air Quality
and Health Benefits
Modelina: Relative
Benefits Derived from
Minnesota combined cycle (IGCC) units with two comparable Operation of the
Public Utilities supercritical pulverized coal (SCPC) units.
Commission
The 2005 analysis used REMSAD to model Hg and
PM air quality changes, and BenMAP to estimate
and value the PM-related health effects. BenMAP
systematically analyzes the health and economic
benefits of air pollution control policy scenarios.
MEP-I/II IGCCPower
Station. December.
Installing IGCC
The annual value of the
2005-
technology would one year of reduced 2012
reduce annual health effects was
emissions by: estimated to be $99
2,600 tons of SO2, million nationally, with
. 600 tons of NOx, ' $24 mllllon ฐccurnn9
, within Minnesota.
12 pounds of Hg.
In 2012, the IGCC
units would avoid:
12 premature
deaths nationally,
20 heart attacks
(infarctions),
eight new cases
of chronic
bronchitis, and
200,000 work loss
days.
The study also
quantified estimates
of other health
effects ranging from
hospital admissions
to asthma attacks.
P
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 172
-------
Chapter 4: Air Pollution, Greenhouse Gas, Air Quality & Health Benefits of Clean Energy Programs
Case Studies
Key Benefits Findings, Results and Activities
State/Region
and Name of
Program
Summary of Policy/Program and Analysis
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Texas: Energy
Efficiency/
In 2001, the 77th Texas Legislature established the
Texas Emissions Reduction Plan (TERP) with the
Renewable enactment of Senate Bill 5, which required the Texas
Energy Impact Commission on Environmental Quality (TCEQ) to
in the Texas promote EE/RE to meet ambient air quality standards
Emissions and to develop a methodology for computing
Reduction Plan emission reductions for State Implementation Plans.
To improve Texas air quality, TERP adopted the goal
Wisconsin:
Haberl et al 2007
Energy Efficiency/
Renewable Energy
Impact in the Texas
Emissions Reduction
Plan (TERP): Volume
1 - Summary Report.
Prepared for the
of implementing cost-effective EE/RE measures Texas Commission on
to reduce electric consumption by 5% per year Environmental Quality
for five years, beginning in 2002, using a variety (TCEQ). August.
of mandatory programs and voluntary financial
incentive programs in non-attainment and affected
counties.
An analysis was performed with data from the TCEQ
and EPA, including eGRID, to estimate the energy
savings and NOX reductions from energy code
compliance in new residential construction.
Focus on Enemy Public Benefits Evaluation -
Focus on Energy Semiannual Summary Report. Prepared by PA
Program Government Services for the Wisconsin DOA.
September 27 2006.
revised in December.
Erickson et al. 2004.
Erickson. J.. C. Best.
D. Sumi. B. Ward.
B. Zent, and K.
Hausker. Estimating
Seasonal and Peak
Environmental
Emission Factors -
Final Report. Prepared
by PAGovernment
Services for the
Wisconsin DOA. May.
Department of
Administration.
State of Wisconsin.
2005. Focus on
Energy Public
Benefits Evaluation -
Semiannual Summary
Report. Prepared
by PAGovernment
Services for the
Wisconsin DOA.
September.
Annual energy
savings in 2006
amounted to:
498,582 MWh of
electricity and
NOX emissions
reduced by:
346 tons per year
in 2004
361 tons per year
576,680 BTUs of in 2006
natural gas
From 2001-
824 tons per year
in 2007
1,416 tons per
year in 2012
2,121 tons per
year 2013
These programs
Add nearly $1 billion
2006, Wisconsin have displaced in value to Wisconsin's
estimated that its annual emissions gross state product
programs saved from power
1 billion kWhs plants and utility
and nearly 50
million therms
in annual energy
customers by:
5.8 million pounds
nf NO
consumption - x
2.6 billion pounds
of CO2
11.4 million
pounds of SOx
46 pounds of
mercury
2002-
2013
2001
2011
RandP
RandP
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 173
-------
Chapter 5: Economic Benefits of Clean Energy Programs
Case Studies
Key Benefits Findings, Results and Activities
State/Region
and Name of
Program
Summary of Policy/Program and Analysis
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
California:
This study analyzes benefits of concentrating solar
Economic, power (CSP) for CA for two deployment scenarios:
Energy, and
Environmental
Benefits of
Concentrating
Solar Power
$7B and $13B invested (2100 MW and 4,000 MW)
from 2008-2020. It emphasized in-state impact of
employment created from manufacture, installation,
and operation of CSP plants.
CSP performance and cost analyzed with Excelergy.
Displaced emissions estimated with emission factors
from California Air Resources Board. Macroeconomic
effects modeled with RIMS II.
Connecticut: In 2004, Connecticut analyzed the economic
Steps in a impact of oil and natural gas conservation policies
Macroeconomic in Connecticut. The state wanted to explore the
Impact impacts of fully funding a program between 2005
Analysis: Oil and 2020 to increase the efficiency of oil and natural
and Natural Gas
Conservation
rOllClGS
gas for residential, commercial, and indus-trial users.
Using the REMI Policy Insight model, their analysis
showed economic benefits to the state from the
increased investment in efficiency and that the
natural gas conservation efforts contributed more
than the oil programs to the overall benefits of the
program.
Note: the expected emissions benefits of these
oil and gas policies is discussed above under the
Chapter 4 case studies
Stoddard. L. J.
Abiecunas. and R.
O'Connell. 2006.
Economic. Energy.
and Environmental
Benefits of
Concentrating Solar
Power in California
Prepared by Black &
Veatch for U.S. DOE
National Renewable
Energy Laboratory.
April.
REMI. 2004.
Economic Impact
of Oil and Natural
Gas Conservation
Policies. Prepared for
U.S. Environmental
Protection Agency
and the State
of Connecticut
November
CSP scenarios avoid
Each 100 MW of
Each dollar spent on
2008
between 8%-18% CSP avoids (per CSP yields direct and 2020
of peak electricity year): indirect impact of $1.40
demand growth by to GSP
74 tons of NO
OC\OC\ X
emissions Each 100 MW of CSP
4000 MW of CSP ... ,,,~~ yields 94 permanent
.,*,_... 2.6 tons of VOCs y K
avoid $60M per jobs
year of natural gas 191,000 tons of
costs in CA CO,
d.
Benefits to the State 1 2005-
Employment (Average 2020
Annual Increase)*:
2,092
Output (Mil '96$) :
3,094.90
GSP(Mil'96$): 2,033.01
Real Disposable
Personal Income (Mil
'96$): 1,749.42
State Revenues (Mil
'01$): 382.13
*Employment is the
average annual increase
from the baseline.
Employment is not
cumulative and is based
on output growth.
P
P
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 174
-------
Chapter 5: Economic Benefits of Clean Energy Programs
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Emissions, Air
Energy Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Georgia:
Quantifying the
This study analyzes benefits of energy efficiency
improvements from 2005-2015 for three investment
Energy System scenarios: minimally, moderately, and very
Benefits of Clean aggressive. Analysis included four main parts:
Energy Policies collect GA energy profile data; estimate EE potential;
Iowa: The
Economic
Impact of Energy
estimate benefits; and review policy options to
achieve EE potential.
EE potential was modeled with ICF's EEPM. Direct
energy cost savings were modeled with ICF's IPM.
Macroeconomic effects modeled with Georgia
Economic Modeling System (GEMS). Public health
effects estimated with EPA's COBRA model.
This study examined the long-term economic
development implications of energy efficiency
(EE) programs, energy pricing/ cost changes, and
Efficiency renewable energy (RE) (biomass and wind power)
Programs and
Renewable
Power
from 1995-2015.
Program cost and savings, including RE cost and
productivity, estimated using program survey data.
Macroeconomic effects (in terms of business output,
personal income and employment) were modeled
with REMI. Results were distinguished by year over
Jensen. V.. and E.
Lounsbury. 2005.
Assessment of Energy
Efficiency Potential
in Georgia. Prepared
for the Georgia
Environmental
Facilities Authority by
ICF Consulting. May.
Weisbrod. G.. K.
Polenske, T. Lynch,
and X. Lin. 1995.
The Economic
Impact of Energy
Efficiency Programs
and Renewable
Power for Iowa: Final
Report. Economic
Development
RACAarrh (~,rm in
a twenty-year period, and broken down by business tr~
> > r y Boston. MA.
tvpe
Pld^ci'ml-icir
Avoided generation
All estimates versus
1.6 - 2.8 job impact per
2005-
in 2010 ranges from 2010 baseline. $1M net benefit 2015
' . ro pmisdnn - r.a^arot-a 1 Rnn /ionn
^-Wp cnuooiuii
Regional wholesale reduced 0.6%-
electricity costs 24%
reduced by U.ID/O
3.9% by 2015
Reduce peak
demand 1.7% 6.1%
h\/ ?D1R
uy CL\Jฑ~J
From 2001-
SO2 emissions
reduced 0.2%-
1.3%
NOX emissions
reduced 0.3%-
1.9%
These programs
- \_JCI ICI dkC J.vJV-'V-' ^^WW
net jobs by 2015
Increase personal
income $48 - $157M
by 2015
REMI model forecasts
1995
P
RandP
2006, Wisconsin have displaced indicate that, in Iowa 2015
estimated that its annual emissions over the 1995-2015
programs saved from power | period, EE can lead to:
1 billion kWhs
and nearly 50
million therms
in annual energy
consumption
plants and utility 25 job-years for every
customers by: $1M invested
5.8 million pounds * $150 of disposable
of J^Q income for every $1
invested
2.6 billion pounds
of cQ Biomass can lead to:
84 job-years per $1M
11.4 million invested
pounds of S0x . $1.45 disposable
46 pounds of income per dollar
mercury
invested
Wind can lead to:
2.5 job-years per $1M
invested
' P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 175
-------
Chapter 5: Economic Benefits of Clean Energy Programs
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Emissions, Air
Energy Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Illinois: The
This 2005 study analyzes benefits of Illinois
Economic and Sustainable Energy Plan RE supplying 8% of
Environmental generation by 2012, 16% by 2020; Reduce load 16%
Impacts of by 2020 with EE; 1570 MW of CHP by 2020; 2000
Clean Energy MW of IGCC by 2020. Measures analyzed separately
Development
and collectively.
Emission savings assume displacement coal-fired
electricity, and estimated with emission factors and
other EIA, EPA, DOE, and EPRI data. Macroeconomic
effects modeled with ILREIM.
Massachusetts: The Massachusetts Division of Energy Resources
Summary of (DOER) produces an annual report analyzing the
Economic impacts of ratepayer-based energy efficiency
Impacts of 2002
Massachusetts
programs in the state. The 2004 report is a
retrospective analysis, using the REMI Policy Insight
Energy Efficiency model, of the macroeconomic effects of investments
Program in energy efficiency (EE) made in 2002. DOER also
Activities
used expenditure and savings data in combination
with the Energy 2020 model to project the lifetime
energy savings of the 2002 program activities.
Bournakis. A.. G.
Hewings. J. Cuttica.
and S. Mueller. 2005.
The Economic and
Environmental
Impacts of Clean
Energy Development
in Illinois. Submitted
to the Illinois
Department
of Commerce
and Economic
Opportunity. June.
DOER. 2004.
2002 Energy
Efficiency Activities.
Massachusetts
Division of Energy
Resources.
By 2020, avoid:
The study estimated
. . .... the plan by 2020 would
0.4 million tons dlrectly lead to:
per year (mtpy) . $7 bim { lncrease
nf 9O
ul ฐ x in economic output
0.2 mtpy of NOX
90.1 mtpy of CO2
SI. 5 billion net increase
in personal income
43,000 net new jobs
Combining direct and
indirect benefits would
achieve by 2020:
$18 billion net increase
in economic output
(2.12% increase)
$5.5 billion net increase
in personal income
(1.83% increase)
191,000 net new jobs
(1.85% increase)
Electricity Bill Impacts
2005-
2020
P
2002- R and P
Energy Savings 2020
Total Program Costs:
$138 million
Total Participant Energy
Savings:
- $21.5 million (Ml/year
- Lifetime = $249M
Average Cost for
Conserved Energy:
4.0 C/kWh
Demand Savings
Total Participant
Demand Savings:
$1.2M/year
Systems Impacts
Customer savings
from Lower Wholesale
Energy Clearing Prices:
$19.4 million
Economic Impacts
Number of New Jobs
Created in 2002: 2,093
Disposable Income
from Net Employment
in 2002: $79 million
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 176
-------
Chapter 5: Economic Benefits of Clean Energy Programs
Case Studies
Key Benefits Findings, Results and Activities
State/Region
and Name of
Program
Summary of Policy/Program and Analysis
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
^^^^^^^^^^^H
National:
A 2005 study analyzes benefits of two potential
Redirecting policies: national 20% RPS by 2020, and 20% RPS
America's
Energy: The
Economic and
Consumer
Benefits of Clean
Energy Policies
New Jersey:
with reallocation of $35 billion of fossil fuel and
nuclear subsidies to EE and RE.
This analysis used regional forecast data from EIA
and other sources, along with IMPLAN to estimate
macroeconomic effects.
A 2005 NJ BPU report analyzes benefits of New
Clean Energy Jersey's Clean Energy Program, which includes
Program: 2005 strategies to increase EE and RE. It analyzes annual
Annual Report and lifetime impact of measures installed in 2005. By
2008, program sought to have 6.5% of NJ electricity
provided by RE. By 2012, the program seeks to have
785,000 MWh and 0.6 mcf of natural gas saved per
year from EE.
Navak. N. 2005.
Redirecting
America's Energy:
The Economic and
Consumer Benefits of
Clean Energy Policies.
Prepared by the U.S.
PIRG Education Fund.
February.
NJ BPU. 2005. New
Jersey's Clean Energy
Program: 2005
Annual Report. New
Jersey Board of Public
Utilities. Office of
Clean Energy.
From 2004 to 2005:
20% RPS with
20% RPS with reallocation 2005-
reallocation avoids achieves, by 2020: 2020
by 2020 versus . 154589 net annual new
baseline:: jobs
634M tons of CO2
1.9M tons of SO2
0.8M tons of NOX
Avoided emissions
$6.8B net increase in
wages
$5.9B average annual
net increase in GDP
From 2001-2006 new
2001-
electric energy from 2005 activities, solar owners were 2020
savings and fฐr 2005-2020: estimated to have saved:
renewable energy . ^ 2M tons of CO * $1-1 milliฐn annually in
generation grew by 2 total electricity costs
over 22% * 46,317 tons of SO,
natural gas savings
grew by over 42%
Efficient equipment
installed and
practices put into
effect in 2005 will
continue to save
energy for an
average of 15 years.
The 5-year program
activities resulted
in lifetime energy
savings of:
over 14 million
MWh of electricity
38 million
Dekatherms of
natural gas
788,000 MWh
of renewable
generation.
The programs
have also reduced
electric demand by
450 MW.
21,813 tons of NOX
P
RandP
' P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 177
-------
Chapter 5: Economic Benefits of Clean Energy Programs
Case Studies
Key Benefits Findings, Results and Activities
State/Region
and Name of
Program
Summary of Policy/Program and Analysis
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
New York:
The New York Energy Smart public benefits program,
Analyzing created in 1998, promotes energy efficiency
Macroeconomic across the commercial, industrial, and residential
Benefits of the sectors; advances renewable energy; provides
Energy SmartSM energy services to low income residents of New
Program
Nevada: Using
REPP Labor
York; and conducts research and development.
As part of a comprehensive evaluation process,
NYSERDA produces an annual report detailing the
multiple benefits of E$P on both a retrospective and
prospective basis.
NYSERDA used the REMI Policy Insight model, a
New York Energy
Smart Program
Evaluation and Status
Report. NYSERDA.
Report to the System
Benefits Charge
Advisory Group. May.
2006
New York Energy
Smart Program
Evaluation and Status
macroeconomic model that combines elements of Report: Year Ending
input-output, econometric, and computable general December 31. 2008
equilibrium models, to conduct the analysis. Outlay
and energy savings estimated primarily using actual
program data.
As part of its 1997 restructuring legislation, the
Nevada legislature established an RPS that included
Calculator: The a 5% renewable energy requirement in 2003 and
Case of Nevada's a 15% requirement by 2013. The Nevada American
RPS Federation of Labor-Congress of Industrial
Organizations (AFL-CIO) used the REPP Labor
Calculator to estimate the job diversification effects
of the RPS in 2005.
NYSERDA. Report to
the Systems Benefit
Charge Advisory
Group. Final Report.
March.
Nevada AFL-CIO.
2003. Comments
Submitted to the
Nevada Public
Service Commission:
Procedural Order
No. 3 and Request
for Comments No. 2.
July.
From 1999 - 2005:
Actions to date
The model indicated the
1999-
1040MW avoid (per year): 1 E$P initiatives from 1999- 2017
reduction in peak . 14 miuion tons 2008 have:
demand of co * Created over net 4,900
2 jobs
From 1999-2005: . 3,170 tons of SO2 . Increased personal
The number of . ^75Q tons Qf NQ income by $293 million,
energy service ' x
companies
increased from
, , . _
fewer tnan lu
tn n\/ฃ>r 1 PD
LU Uvtri .LOU
companies
5% renewable
energy requirement
in 2003
15% requirement
by 2013
GSP by $644 million
Total output by $1
billion
Projecting to 2020, E$P is
expected to create 86,400
net job years.
From 2008-2017 actions
to date yield (per year):
Average of 4,100 jobs
$182M labor income
$244M output
From 2003-2013, the RPS
would create:
27229 total, direct FTE
jobs
Of which, 19,138 are
manufacturing jobs and
8,092 would be
installation, O&Mjobs
*excludes indirect or
induced effects
2003
2013
RandP
P
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy
178
-------
Chapter 5: Economic Benefits of Clean Energy Programs
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Emissions, Air
Energy Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Oregon:
Economic
Impacts of
Energy Tax
Credit Programs:
The Oregon Department of Energy asked
ECONorthwest to estimate the economic effects
of the Business Energy Tax Credit (BETC) and
Residential Energy Tax Credit (RETC) programs.
These effects include impacts on employment,
BETC/RETC output, and wages as well as tax revenue in
Southwest: The
New Mother
Lode: The
Potential for
More Efficient
Electricity Use
Oregon that resulted from 2006 tax credits and
the subsequent spending on measures and labor
that these credits create. ECONorthwest also
isolated the economic impacts of energy efficiency
improvements (i.e., energy savings) that were
realized in 2006 in order to estimate the benefits to
the economy that accumulate in future years.
They used IMPLAN to model the macroeconomic
effects.
The Southwest Energy Efficiency Project (SWEEP)
analyzes benefits from $9B invested in EE in homes
and businesses in the Southwest from 2003-2020
by comparing a BAU scenario to a "High Efficiency"
scenario. "High efficiency" assumes widespread
adoption of cost-effective, commercially available EE
measures that would reduce electricity consumption
by 18% by 2010 and 33% by 2020.
Residential and commercial cost-effective energy
savings modeled with DOE-2.2. Industrial cost-
effective energy savings potential modeled with LIEF.
Energy cost savings and avoided emissions modeled
with NEMS. Macroeconomic effects modeled with
IMPLAN.
Grover. S. 2007.
Economic Impacts
of Oregon Energy
Tax Credit Programs
in 2006 (BETC/
RETC). Prepared by
ECONorthwest for the
Oregon Department
of Energy. May.
SWEEP. 2002. The
New Mother Lode:
The Potential for More
Efficient Electricity
Use in the Southwest.
Southwest Energy
Efficiency Project.
Report for the Hewlett
Foundation Enercrv
Series November
Oregon
commercial and
residential energy
costs decreased by
$46 million
By 2020:
Avoids $10.6B
capacity
investment (thirty-
five 500 MW plants)
Avoids $25B
electricity supply
costs per year by
2020
Avoids $2.4B end-
use natural gas cost
per year by 2020
By 2020:
Reduces CO2
emissions by 26%
Reduces SO,
emissions by 4%
Reduces NOX
emissions by 5%
The net impacts of the
2006-
tax credits in Oregon for 2021
the year 2006 were an
increase in:
Gross state product of
more than $142 million
Jobs by 1,240
Tax revenue of nearly
$10 million
Oregon wages by $18.6
million
Continued energy
savings support the
following annual
economic impacts in
future years:
Increase in Oregon's
economic output by
$93 million
Continued net impact
of 889 new jobs
Additional state and
local tax revenues of
$10 million
Increase regional
employment by 0.45%
(58,400) FTE jobs
per year versus 2020
baseline
Increase salary income
by $1.34B per year
versus 2020 baseline
2003-
2020
RandP
P
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 179
-------
Chapter 5: Economic Benefits of Clean Energy Programs
Case Studies
Key Benefits Findings, Results and Activities
State/Region
and Name of
Program
Summary of Policy/Program and Analysis
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Utah: Using
JEDI: The Case
of Wind Power in
Wind power has been proposed in Utah as a way to
diversify the state's electricity generation. In 2006
Utah State University used JEDI to inform decision
Utah County makers about the likely impact of five wind capacity
scenarios: 5 MW, 10 MW, 14.7 MW, 20 MW, and 25
MW. This report quantifies the potential economic
opportunities created by wind development,
including projections for the 14.7-MW project in
Spanish Fork Canyon, for Utah County. It uses
economic and demographic information from three
sources: (1) the Economic Development Corporation
of Utah (EDCU); (2) IMPLAN multipliers for Utah
county supplied by NREL; and (3) two local wind
developers.
Wisconsin: Wisconsin's Focus on Energy Program advances cost
Focus on Energy effective energy efficiency and renewable energy
Program projects in the state through information, training,
enerav audits, assistance and financial incentives.
The Wisconsin Department of Administration
conducted an evaluation of the economic impacts
of the Focus on Energy Program from its inception
in 2002 through 2026. The analysis involved
documentation and extrapolation of the net direct
effects of the program; application of a regional
economic model; and analysis of the implications.
The results indicate that the Focus on Energy
Program provides net benefits to the State of
Wisconsin.
Monaha. N.. E.
Stafford, and C.
Hartman. 2006.
An Analysis of the
Economic Impact
on Utah County.
Utah from the
Development of
Wind Power Plants.
Renewable Energy
for Rural Economic
Development. Utah
State University. DOE/
GO-102006-2516.
May.
Wisconsin
Department of
Administration. 2007.
Division of Energy.
Focus on Energy
Public Benefits
Evaluation. Economic
Development
Benefits: FY07
Economic Impacts
Report. Final: February
2^2QQZ
If the Spanish Fork
Not
project (14.7 MW) were specifed
built it would produce
(using 2005 dollar
values):
46 total new jobs
$1.2 million in wage
earnings
$4.2 million in
economic output
during the construction
phase of the project
Between 2002 and 2026, 2002 -
the Focus on Energy 2026
Program is expected to:
create more than
ฃD ODD inl-i v^arc'
DU,UUU JUU~yccub,
generate sales for
Wisconsin businesses
of more than eight
billion dollars;
increase value added or
gross state product by
more than five billion
dollars;
increase disposable
income for residents by
more than four billion
P
dollars.
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 180
-------
Additional Studies and Programs that Highlight the Benefits of Clean Energy
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Emissions, Air
Energy Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
California: The
Economics of
Solar Power
The Million Solar Roofs initiative seeks to install
3000 MW of solar on CA roofs by the end of 2016.
This analysis covers retrofit and new construction
applications between 2007-2016 and estimates the
multiple benefits of the initiative.
Infrastructure and emission savings based on
E3 Avoided Cost model. Primary analysis was
performed with Million Solar Systems Model, based
on solar market data from CEC and CPUC.
Massachusetts: This study analyzes retrospectively the benefits
The Public of EE in MA from 1977-1997 and projects future
Benefit of Energy benefits through 2015. Study does not establish a
Efficiency to link between actual government EE programs and
the State of changes in EE.
assac use s ^ uses an econometric model. Changes in energy
intensity used to approximate efficiency changes by
Midwest:
controlling for sector composition, energy prices,
new capital, and climate.
This study analyzes benefits of implementing
Job Jolt: The the Repowering the Midwest Clean Energy
Economic Development Plan for a 10-state region in the
Impacts of Midwest that includes reducing electricity demand
Repowering the by 28% by 2020 with EE, and diversifying towards RE
Midwest: The
Clean Energy
Development
Plan for the
Heartland
National: 2002
Energy Efficiency
and CHP generation over a 20-year period.
The analysis is performed with Census and other
data, and econometric I-O models developed by
REAL at the University of Illinois.
This study analyzes benefits of $138 million of
ratepayer-based EE investments during 2002 and
Activities: A cumulative EE investments from 1998-2002. It
Report by analyzes annual and lifetime benefits to participants
the Division and all consumers.
of Energy
Resources
The energy cost savings, energy system benefits
and emission savings estimated with actual program
dcitci, loU~JNt dcitci, otrisr dcitci, uuts tnsrcfy cXjcX),
and a bid-stack model. Macroeconomic effects are
modeled with REMI.
Cinnamon. B.. T
Beach. M. Huskins.
and M. McClintock.
2005. The Economics
of Solar Power for
California. White
Paper Aucrust
Bernstein, M.,
R. Lempert. D.
Loughram. and D.
Ortiz. 2002a. The
Public Benefit of
Energy Efficiency
to the State of
Massachusetts
Prepared by RAND
Science and
Technology.
Hewings and Yanai.
2002. Job Jolt: The
Economic Impacts
of Repowering the
Midwest: The Clean
Energy Development
Plan for the Heartland.
Commonwealth
of Massachusetts
Office of Consumer
Affairs and Business
Regulation. 2004.
2002 Energy
Efficiency Activities: A
Report by the Division
of Energy Resources.
Summer
Avoid $71M
Avoid $5,526M in
Additional $0.50
2007
capacity emission costs, economic activity in CA 2016
infrastructure costs including NOX per $1 invested
(3,000 MW of peak
capacity)
$19.4M savings
from 1998-2002
andCO2 ... ___ . , . .
40 FTE jobs in CA per
MW
P
In 1997 past energy From 1977-1997 EE 1 1977 1 R and P
efficiency actions produced $1,644 - 1997; and
resulted in a $2,562 in per capita through
reduction of: GSP 2015
2.0M tons of CO2 $323 - $2,322 additional
i , r~.r~.r~. . t o^ Per capita gains by
11,000 tons of SO2 9Q15
4,000 tons of NOX
(Versus 1997
baseline)
2002 emission
By 2020:
2002-
2020
Over 200,000 net new
jobs
$19.4B increase in
regional economic
output
In 2002:
reductions: . 1J78 new jobs
($5.9M for 2002 . 394 tons SO2 . $139M m Gsp
only) due to . 135 tons NO ^n,, A.
lower wholesale x $79M disposable
electricity prices 161,205 tons CO2; income
0.5% (48 MW) peak ufetime effect of Lifetime effect of 2002
H^manH r^Hi irtinn 2002aCtlOnS: actions
UcllldllU 1 cUUdlUI I
in 2002.
5,516 tons SO2 . 315 permanent jobs,
. 1,890 tons NOX . $22M Gsp
2,256,870 tons . $15M in income
C02
1998-
2002
P
R
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 181
-------
Additional Studies and Programs that Highlight the Benefits of Clean Energy
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
National: The
This study analyzes labor requirements for
Work that Goes renewable energy deployment in the United States.
Into Renewable Labor estimates from construction, installation,
Energy and O&M only account for direct effects - indirect
multiplier effects no examined. The study is not
specific to any particular state and used survey
information, not a model. Authors collected primary
employment data from companies in the RE and
coal sectors. It accounts for jobs in manufacturing,
National:
Ancillary Benefits
of Reduced Air
Pollution in the
United States
from Moderate
Greenhouse
Gas Mitigation
Policies in the
Electric Sector
New England:
Electric Energy
Efficiency and
Renewable
Energy: An
transport, delivery, construction, installation, and
O&M and includes a comparison with coal power.
This study analyzes benefits of GHG and criteria
pollutant mitigation, including the value of health
Singh and Fehrs .
2001. The Work that
Goes Into Renewable
Energy. November.
Burtraw et al. 2001.
Ancillary Benefits of
impacts from air quality changes. It analyzes various Reduced Air Pollution
carbon-tax scenarios from 2000-2010. in the United States
The analysis used the Haiku electricity model to
simulate effects on retirement and system dispatch.
Emission changes were translated into health effects
with damage functions and the TAP atmospheric
transport model. Concentration-Response functions
were used to estimate health endpoints.
This study aAnalyzes benefits of EE and RE in
New England from Public Benefits Funds and RPS
programs. It assumes that current policies change
only as planned, through 2010, and does not
cover unplanned scenarios. Authors used actual
Assessment of and estimated data on program expenditures and
Existing Policies savings. Air quality and emission benefits were
from Moderate
Greenhouse Gas
Mitigation Policies in
the Electric Sector.
December.
Sedano et al. 2005.
Electric Energy
Efficiency and
Renewable Energy:
An Assessment of
Existing Policies and
Prospects for the
and Prospects for estimated with OTC's Emission Reduction Workbook Future. May.
the Future
Pennsylvania:
Economic
Impact of
Renewable
Energy
and macroeconomic effects were modeled with
IMPLAN.
This study analyzes benefits of implementing a 10%
RPS in PA over the period 2006-2025, which would
require $4.68 billion direct investment. A statewide
renewable energy supply curve was created to
determine the least-cost portfolio. Authors used a
simple linear model with publicly available data, and
Pletka, R. 2004.
Economic Impact of
Renewable Energy
in Pennsylvania.
Prepared by Black
& Veatch for The
the BEA's RIMS II model to estimate macroeconomic Heinz Endowments
effects.
and Community
Foundation for the
Alleghenies. March.
In 2004, EE reduced
peak demand by
NOX related health
benefits in 2010
range from $315 -
$408M
NOX related health
benefits per ton of
carbon emissions
reduced, range
from $7.5 - $13.2
dollars
From 2000 - 2010,
avoid:
Job effects:
35.5 person-years per
MW of solar
4.8 person-years per
MW of wind
3.8-21.8 person-years
per MW of biomass co-
firing
5.7 person-years per
$1M solar or wind cost
over 10 years
From 2000 - 2010, net
increase of:
1,421 MW ^ ^ ?M ton^ (fio/) . $6 1B economic output
of CO $1.04M wage income
. 34,200 tons of SO, ' 28'190 Jฐb years
22,039 tons of
NOX
Over 2006-2015 period:
2001
2000-
2010
2005-
2010
2006-
Increase output $10.1B 2025
Increase earnings $2.8B
Create 85,000 jobs
R
P
P
P
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 182
-------
Additional Studies and Programs that Highlight the Benefits of Clean Energy
State/Region
and Name of
Program
Case Studies
Summary of Policy/Program and Analysis
Key Benefits Findings, Results and Activities
Energy
Emissions, Air
Quality, and Health
Economic
Time
Period for
Analysis
Type of
Analysis*
Texas: Increasing
This study analyzes benefits of increasing Texas'
the Renewable current RPS (requiring 2.7% of sales from new
Energy Standard: renewable energy by 2009) to a requirement of 20%
Economic and renewable energy by 2020. It also analyzes a more
Employment modest increase to about 8% renewable energy by
Benefits 2025.
Impacts on electricity and natural gas prices and
consumer energy bills were examined using the
Department of Energy's National Energy Modeling
System (NEMS) model. Macroeconomic impacts
were quantified using IMPLAN. Expenditure
breakdown and local share data for wind projects
were based on NREL's Jobs and Economic
Development Impacts (JEDI) model
Washington: This study analyzes the benefits of an RPS that
The Washington would support 1,300 average megawatts (avgMW)
Clean Energy of renewable sources by 2025, along with 1,000
Initiative: Effects avgMW of cost-effective energy efficiency from
of 1-937 on 2010-2025. The analysis compares the clean energy
Consumers, initiative with a reference case in which no further
Jobs and the
Economy
energy efficiency and renewable energy investments
are made after 2009.
Effects on electricity rates, total resource costs, and
consumer electricity bills were examined using a
spreadsheet model. Macroeconomic impacts were
analyzed using IMPLAN. Expenditure breakdown
data for construction, O&M of renewable plants was
based on a variety of sources, including state and
federal agencies, renewable developers, utilities, and
NREL's Jobs and Economic Development Impacts
Deyette and Clemmer.
2005. Increasing the
Renewable Energy
Standard: Economic
and Employment
Benefits.
Deyette and
Clemmer. 2006. The
Washington Clean
Energy Initiative:
Effects of 1-937 on
Consumers. Jobs and
the Economy.
By 2025, the 20% RPS
achieves:
9% reduction in
By 2025, the 20%
RPS avoids:
?fl millinn mptrir
average electricity tons of CO2
prices emissions
3% reduction in
natural gas prices
Residential solar
heating systems
that offset 390 MW
of peak capacity
The set of efficiency
By 2020, the 20% RPS
2005-
achieves: 2025
$950M additional
income
$440M increase in GSP
24,650 net new jobs
(2.8 times more jobs
than with fossil fuels)
By 2025, the By 2025, the initiative 2010-
measures developed initiative avoids: achieves: 2025
under the initiative ._ .... . t-ivon jj-i- i
, . 4.6 million metric S138M additional
achieve: ,
tons Of CO inrnmp
An average savings emissions
of $0.54 cents/kWh
due to avoided T&D
Avoided
construction of six
natural gas power
plants, operating at
an average capacity
of 165 MW each.
$148M increase in GSP
$30M in income to
rural landowners
1,230 net new jobs
in the year 2025 (2.6
times more jobs than
would be created using
fossil fuels)
(JEDI) model.
P
P
* P = Prospective; R= Retrospective
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 183
-------
This page is intentionally left blank.
APPENDIX A | Assessing the Multiple Benefits of Clean Energy 184
-------
APPENDIX B
Tools and Models Referenced
in Each Chapter
U
o
Q
CHAPTER ONE
Introduction
CHAPTER TWO
Potential Energy Impacts of Clean Energy
CHAPTER THREE
Electric System Benefits of Clean Energy
CHAPTER FOUR
Air Quality Benefits of Clean Energy
CHAPTER FIVE
Economic Benefits of Clean Energy
APPENDIX A
Catalogue of Clean Energy Case Studies
APPENDIX B
Tools and Models Referenced in Each Chapter
APPENDIX B | Assessing the Multiple Benefits of Clean Energy 185
-------
TOOLS AND MODELS REFERENCED IN
CHAPTER 2
EXAMPLES OF AVAILABLE TOOLS FOR
ESTIMATING DIRECT ENERGY IMPACTS
Internet-Based Methods:
eCalc
WEB SITE: http://ecalc.tamu.edu/
EPA Energy Savings Calculators
WEB SITE: http://www.energystar.gov/purchasing
ENERGY STAR Roofing Comparison Calculator
WEB SITE: http://www.roofcalc.eom/default.aspx
- ENERGY STAR Target Finder
WEB SITE: http://www.energystar.gov/index.cfm?c=new_
bldg_design. bus_target_finder
- ENERGY STAR Portfolio Manager
WEB SITE: https://www.energystar.gov/benchmark
PVWatts
WEB SITE: http://rredc.nrel.gov/solar/codes_algs/
PVWATTS/versionl/
Spreadsheet-Based Methods:
EMD International WindPro
WEB SITE: http://www.emd.dk/WindPRO/Introduction/
RETScreen Clean Energy Project Analysis Software
WEB SITE: http://www.retscreen.net/ang/home.php
Integral Analytics: DSMore
WEB SITE: http://www.integralanalytics.eom/dsmore.php
Software Methods:
fChart and PV-fChart
WEB SITE: http://www.fchart.eom/index.shtml
EQuest
WEB SITE: http://www.doe2.eom/equest/
ENERGY-10
WEB SITE: http://www.nrel.gOV/buildings/energylO.html
DOE-2
WEB SITE: http://doe2.cont/DOE2/index.html
EXAMPLES OF SOPHISTICATED SUPPLY
FORECASTING MODELS
Electricity Dispatch:
PROSYM
WEB SITE: http://wwwl.ventyx.eom/analytics/
market-analytics.asp
GE-MAPS
WEB SITE: http://www.gepower.com/prod_serv/
products/utility_software/en/ge_maps/index.htm
PROMOD
WEB SITE: http://www.ventyx.com/analytics/
promod.asp
- MIDAS
WEB SITE: http://www.ventyx.cont/advisory/
horizons-interactive, asp
Capacity Expansion or Planning:
NEMS
WEB SITE: http://www.eia.doe.gov/oiaf/aeo/
overview/index.html
IPM*
WEB SITE: http://www.icfi.eom/Markets/Energy/
energy-modeling.asp#2
ENERGY 2020
WEB SITE: http://www.energy2020.eom/
Long range Energy Alternatives Planning System
(LEAP)
WEB SITE: http://www.energycommunity.org/
default.asp?action=47
Strategist
WEB SITE: http://wwwl.ventyx.eom/analytics/
strategist, asp
Plexos
WEB SITE: http://www.energyexentplar.cont/
APPENDIX B | Assessing the Multiple Benefits of Clean Energy 186
-------
EGEAS
WEB SITE: http://my.epri.com/portal/server.pt?space
=CommunityPage&cached= true&parentname=0
bjMgr&parentid=2&control=SetCommunity&Co
mmunityID=221&PageIDqueryComId=0
AURORAxmp
WEB SITE: http://www.epis.com/aurora_xmp/
MARKAL-MACRO
WEB SITE: http://www.etsap.org/Tools/MARKAL.
htm.
1 Ventyx System Optimizer
WEB SITE: http://www.ventyx.cotn/analytics/systetn-
optimizer.asp
TOOLS AND MODELS REFERENCED IN
CHAPTER 3
DISPATCH MODELS AVAILABLE FOR STATES
EnerPrise Market Analytics (powered by PRO-
SYM) supported by Ventyx.
WEB SITE: http://wwwl.ventyx.cotn/analytics/
market-analytics.asp
Multi-Area Production Simulation (MAPS)
developed and supported by GE Energy and sup-
ported by other contractors.
WEB SITE: http://www.gepower.com/prod_serv/
products/utility_software/en/ge_maps/index.htm
* Plexos for Power Systems owned by Energy
Exemplar.
WEB SITE: (http://www.energyexetnplar.cotn)
PowerBase Suite (including PROMOD IV*) sup-
ported by Ventyx.
WEB SITE: http://wwwl.ventyx.cotn/analytics/
promod.asp
ELECTRIC SECTOR-ONLY CAPACITY
EXPANSION MODELS
IPM* developed and supported by ICF
International.
WEB SITES:
http://www. icfi. com/Markets/Energy/energy-
modeling.asp#2
http://www.icfi.com/markets/energy/docjlles/
iptnglobal.pdf
PowerBase Suite (including Strategist*) supported
by Ventyx.
WEB SITE: http://wwwl.ventyx.cotn/analytics/
strategist, asp
WHOLE ENERGY-ECONOMY SYSTEM
PLANNING MODELS
Energy system-wide models with electricity
sector capacity expansion:
U.S. DOE National Energy Modeling System
(NEMS)
WEB SITE: http://www.eia.doe.gov/oiaf/aeo/
overview/
APPENDIX B | Assessing the Multiple Benefits of Clean Energy 187
-------
MARKet ALlocation (MARKAL) Model
WEB SITE: http://www.etsap.org/markal/main.html
' Energy 2020
WEB SITE: http://www.energy2020.com/
Specialized proprietary models of the T&D
system's operation:
PowerWorld Corporation's power systems simula-
tion package
WEB SITE: http://www.powerworld.com/
Siemens (PSS*E) probabilistic analyses and dynam-
ics modeling
WEB SITE: https://www.energy.siemens.com/
cms/00000031/en/ueberuns/organizati/services/
siemenspti/softwareso/Pages/psse_1439533.aspx
TOOLS AND MODELS REFERENCED IN
CHAPTER 4
TOOLS TO HELP STATE AND LOCAL
GOVERNMENTS DEVELOP GHG AND
CRITERIA AIR POLLUTANT EMISSION
INVENTORIES
EPAs State Inventory Tool (SIT)
WEB SITE: http://www.epa.gov/climatechange/
emissions/state_guidance.html
Clean Air and Climate Protection Software Tool
(CACPS)
WEB SITE: http://www.cacpsoftware.org/
TOOLS STATES CAN USE TO HELP DEVELOP
BOTTOM-UP GHG AND CRITERIA AIR
POLLUTANT INVENTORIES
For GHG inventories:
Portfolio Manager
WEB SITE: http://www.energystar.gov/
index, cfm ?c=evaluate_performance.
bus_portfoliomanager_carbon
For criteria air pollutant inventories:
Point Sources: Landfill Gas Emissions Model
WEB SITE: http://www.epa.gov/ttn/catc/dirl/
Iandgem-v302-guide.pdf
Mobile sources:
MOBILE6
WEB SITE: http://www.epa.gov/OMS/m6.htm
NON ROAD 2005
WEB SITE: http://www.epa.gov/oms/nonrdmdl.htm
Motor Vehicle Emission Simulator (MOVES)
WEB SITE: http://www.epa.gov/otaq/models/moves/
index.htm
DATA SOURCES AND ADDITIONAL
RESOURCES FOR TOP-DOWN AND BOTTOM-
UP INVENTORIES
National Emissions Inventory (NEI)
WEB SITE: http://www.epa.gov/ttnchiel/trends/
APPENDIX B | Assessing the Multiple Benefits of Clean Energy 188
-------
eGRID
WEB SITE: http://www.epa.gov/cleanenergy/energy-
resources/egrid/index.html
Emissions Collection and Monitoring Plan System
(ECMPS)
WEB SITE: http://www.epa.gov/airmarkets/business/
* WRI Climate Analysis Indicators Tool:
WEB SITE: http://cait.wri.org/
State Agencies and Universities
EPA State GHG Inventories
Local GHG Inventories
TOOLS FOR FORECASTING FUTURE
EMISSIONS
EPA EIIP Technical Report Series, Volume X:
Emissions Projections.
WEB SITE: http://www.epa.g0v/ttn/chief/eiip/
techreport/volumel 0/xOl.pdf
EPA State GHG Projection Tool.
WEB SITE: http://www.epa.gov/climatechange/wycd/
stateandlocalgov/analyticaltools.html
* The Clean Air and Climate Protection Software
Tool.
WEB SITE: http://www.cacpsoftware.org/
BASIC AND SOPHISTICATED APPROACHES
FOR QUANTIFYING AIR POLLUTANT AND
GHG EMISSION EFFECTS OF CLEAN ENERGY
INITIATIVES
Basic Approaches:
eCalc
WEB SITE: http://www.ecalc.eom/calculator/
scientific/
- OTC Workbook
CACPS
WEB SITE: http://www.cacpsoftware.org/
Sophisticated Approaches:
Electric Dispatch
PROSYM
WEB SITE: http://www.ventyx.com/analytics/
market-analytics.asp
GE-MAPS
WEB SITE: http://www.gepower.com/prod_serv/
products/utility_software/en/ge_maps/index.htm
PROMOD
WEB SITE: http://www.ventyx.eom/analytics/
promod.asp
Capacity Expansion or Planning
NEMS
WEB SITE: http://www.eia.doe.gov/oiaf/aeo/
overview/index.html
IPM*
WEB SITE: http://www.icfi.com/Markets/Energy/
energy-modeling.asp#2
ENERGY 2020
WEB SITE: http://www.energy2020.eom/
LEAP
TOOLS FOR QUANTIFYING AIR QUALITY
AND/OR HEALTH IMPACTS
SCRAM
WEB SITE: http://www.epa.gov/ttn/scram/
REMSAD
WEB SITE: http://remsad.saintl.eom
CAMx
WEBSITE: http://www.camx.coni
UAM-V
WEB SITE: http://uamv.saintl.eom
CMAQ
WEB SITE: http://www.epa.gov/AMD/CMAQ/
CMAQscienceDoc.html
APPENDIX B | Assessing the Multiple Benefits of Clean Energy 189
-------
CALPUFF and AERMOD
WEB SITE: http://www.epa.gov/scram001/
dispersion_prefrec. htm
COBRA
WEB SITE: http://epa.g0v/statelocalclimate/
resources/'cobra.html
BenMAP
WEB SITE: http://www.epa.gov/air/benmap/
ASAP
WEB SITE: http://www.epa.gov/ttn/ecas/asap.html
TOOLS AND MODELS REFERENCED IN
CHAPTER 5
SCREENING TOOLS
Job and Economic Development Impact (JEDI) Model
for Wind Projects
WEB SITE: http://www.energyfinder.org/
REPP Labor Calculator
WEB SITE: http://www.repp.org/index.html
RMI Community Energy Opportunity Finder
WEB SITE: http://www.energyfinder.org/
MODELS FOR ESTIMATING MACROECONOMIC
BENEFITS
IMPLAN* input-output model (IMPLAN)
WEB SITE: http://www.implan.com/
RAND
WEB SITE: http://www.rand.org/
REMI Policy Insight model (REMI)
WEBSITE: http://www.remi.com/
Berkeley Energy and Resources model (BEAR)
APPENDIX B | Assessing the Multiple Benefits of Clean Energy 190
------- |